PDA

View Full Version : Tideflex flap valve



boiermoflea
December 1, 2016, 06:06 AM
Hello,

I am having some difficulties in modelling a tideflex valve.
I have received a head discharge curve from manufacturer. Problem is that the head discharge curve changes with the backpressure. (tide)
I cannot model multiple head discharge curve at once, so my problem is which one I should use. Is it worth using the curve drawn with red from the head discharge data received (bellow) as a full submerged situation?
The tide that I'm taking into consideration, is higher than the soffit of the valve even at lowest values.
Here is the head discharge data received:
351
Also here is some info from Innovyze about how to model a tideflex, for whom it may concern:
"Hi,

Unfortunately, there is only the capability to add one of the Head/flow curves to the model. The only useful suggestion would be to pick the most appropriate curve. Usually this would be the curve for the maximum back pressure rating. You may also want to consider which curve is most relevant to the modelling scenarios that you’re undertaking."


"Hello,

The recommended method for modelling tideflex valves is to represent the valve with a user defined head / discharge curve and an oversized flap valve modelled in series.

The user defined head / discharge table is populated from manufacturer’s headloss data and the oversize flap valve is required to prevent reverse flow whilst at the same time generating zero or negligible headloss.

The below schematic shows an example.

352

A 450mm diameter duckbill valve is attached to a 450mm diameter outfall pipe. The invert level of the outfall pipe, the initial level of the user defined control and the flap valve invert are all equal. The ‘dummy’ flap valve has been deliberately oversized to 3m so that negligible headloss is produced. A Cd (flap valve) of 0.1 can be used to keep headloss low."



Thanks

Kristian Ravnkilde
December 5, 2016, 12:36 AM
That looks reasonable. Do you have any sort of information to calibrate it to, such as observed depth in a manhole upstream at high tide?

boiermoflea
December 5, 2016, 01:01 AM
I have such data. But I don't think this will help, since the flow from this manhole is not going entirely on the tideflex valve. Upstream of this tideflex valve there is a bifurcation. The flow will go to a pumps chamber. Just to make things complicated, the pumps used are old, and we don't know exactly the flow rate of the pumps. Not only because the pumps are old and the flow rate is not the same as the one in the manufacturer data, but also the pumps chamber have saline infiltration that can't be quantified at this moment.

Kristian Ravnkilde
December 5, 2016, 01:13 AM
Why am I not surprised to hear that??? Life is never easy..... Your proposed curve is reasonable, so I'd go with that and see if that with your best estimate of pump rate and inflow gives something like the observed performance, then tweak to suit. You may need to do some sensitivity testing to see what range of combinations gives the same answer, and report to the client on that and the associated uncertainty. Then what version you choose to use depends on the situation and what you are proposing to do about it - identifying a "no regrets" solution to whatever the problem is may involve combining elements from the range of sensitivity tests.
Best of luck. And remember my motto: big up the model, diss the data!