TUFLOW Forum

# Model Stability at Weir in Canal

## Recommended Posts

Hi all

I am modelling flood flows in a tidal canal. The canal is approximately 100 metres wide with a Manning's n of 0.025, and an invert level of -3.5mAHD. The model uses a 6 metre grid and 3 second timestep. A 48 metre wide weir is located across the canal, at a level of 0.6mAHD. The 100 year flood level upstream of the weir is 2.0mAHD. Everything is modelled in the 2D domain.

The model runs well and has a low mass balance error of -0.1%, however the flows and water levels in the vicinity of the weir are quite "unstable" when the flood flow starts to increase, particularly just downstream of the weir. Because of the very high conveyance provided by the canal, a small change in water level means a very large change in flow.

Is there a way to make the model results more stable around the weir? I tried halving the timestep, but that did not help. Would the weir be better modelled in 1D?

Thanks

##### Share on other sites

Hi there,

Is the 48m wide weir completely level across its width? Sometimes a slight increase in upstream WL can lead to very sudden increases in discharge over the structure and thus instabilities. You could perhaps stagger the elevation of the 8x6m cells, eg cell 1 = RL 0.57m, 2=0.58m, ... 6=0.64. This may help smooth the sudden increase in flow over the structure.

But yes, modelling the weir in 1D would also be a common approach. Is it a sharp crested weir or broad crested weir? If sharp crested it would be better in 1D. Also check that the d/s boundary is a fair distance downstream, eg, >30-60m. Be sure to test the results against first principles.

Regards,

Paul.

##### Share on other sites

The first thing I can think to ask is how you've constructed your weir in the domain; a thick zline is likely to work that bit more stably than a thin one, due to the U and V zpts being raised as well as the cell centres. In that vein, if your weir is broader than one cell in the direction of flow, try making sure all the cell sides at the edges of your weir are also selected... It may help! Or it may not...

Is your downstream boundary any where near the weir? If so, you might do well to move it a bit further away if possible, especially if it's just an HT.

Finally, if the model is that sensitive to very small changes in water level, you could try running it double precision, which would give TUFLOW that bit more resolution in terms of head and velocity. Shouldn't think it's that, but worth trying just in case!

Alternatively, as you suggest, 1d should be able to cope with the situation described just fine, so if nothing else seems to fix it, try that I suppose.

Let us know how you get on and if you find a fix!

##### Share on other sites

The first thing I can think to ask is how you've constructed your weir in the domain; a thick zline is likely to work that bit more stably than a thin one, due to the U and V zpts being raised as well as the cell centres. In that vein, if your weir is broader than one cell in the direction of flow, try making sure all the cell sides at the edges of your weir are also selected... It may help! Or it may not...

Just re-read what I wrote, and realise that what I say regarding thick and thin z-lines is nonsense! but if you're using a wide z-line, do watch out to make sure your cell sides are selected...

That's all!

##### Share on other sites

Thanks for the responses. To provide some more info to the questions raised:

1. The weir is level along its full width, and is sharp crested.

2. The weir is modelled using a THICK z-line. (Perhaps a THIN line would be better?)

3. I tried using the double precision TUFLOW, but it did not make the results more stable.

4. The boundary of the model is about 1800 metres downstream of the weir.

I will try modelling the weir as a 1D link to see if that improves the stability of the results.

##### Share on other sites

I have modelled the weir as a 1D link, and this has resolved the stability issue around the weir (and gives the expected water level upstream of the weir). Thanks all for your suggestions. However I would still like to understand the reasons why the 2D weir gave such unstable results.

Cheers

## Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

×   Pasted as rich text.   Paste as plain text instead

Only 75 emoji are allowed.

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×

×