open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello forum!

I am currently working with a sisyphe model around a harbour pier and I have a few questions regarding my setup.

What I want to accomplish is to calculate the transport rates along a shore-perpendicular transect outside the harbour entrance. I have a few different harbour layouts and simply want to compare the transport rates between a few different pier designs. I intend to model constant wind and wave conditions.

My current setup is as follows:
Waves are modelled with TOMAWAC until steady state. Thereafter, a Telemac3D-model incorporate the waves using 'WAVE DRIVEN CURRENTS = YES' and calling the last time step in the result file from TOMAWAC. The 3D-model is also coupled with SISYPHE. Of course, it would probably be more accurate to couple all three models, but it is both time demanding and my perception is that this coupling is not possible with constant wave conditions?

I started to doubt my setup when I came across the keyword 'EFFECT OF WAVES' in SISYPHE. Am I not including the orbital wave energy and underestimating the wave effects?

All experts out there, what do you think? Is the setup I'm using sufficient for comparing the transport rates for the different designs or do I need to consider 'EFFECT OF WAVES'?

I'm not sure if this description is clear enough but I'll do my best to clarify if you wish :)

Superthanks in advance!

/Schmirre

Dear Schmirre,
If I understood well your question, I'd say that you need to activate the effect of waves in Sisyphe (EFFECT OF WAVES = YES).
This allows you to account for the wave effects when solving sediment transport rates. Please be aware of the sediment transport formula that your are using and consider to adopt one of the formulas developed for accounting the wave effect, such as Bijker, Bailard or Dibajnia & Watanabe.

I hope it helps,

Best wishes,

Pablo

Hello Schmirre,

You definitely need to set EFFECT OF WAVES = YES (as Pablo said) and then give the wave results file (WAVE FILE='...').

As long as bed evolution does not make significant changes to wave conditions you may get away with 2-way coupling. Also, you need to give some time to the currents to reach a steady state before you start measuring sediment transport. Consider setting up hot-start computations (or a TOM-T3D coupled computation?) and continue from there.

Regards,
Costas

Thank you both Pablo and Costas for your input, I will take your advice!

I believe that the bed evolution in my case have negligible effects on the waves. I am currently running an initialization period on my T3D-model and then simply calling computation continued from my T3D-SIS. That should do it, right?

Yes, that will do fine. As an extra step, you can setup a TOM-T3D coupled computation (for a hot start) so that the wave field you use for subsequent sediment transport computations includes wave-current interaction.
Costas

Yes, that's what I meant by the last post :)

I am currently running my first computation with 'EFFECT OF WAVES = YES' but the computation freezes after only a few time steps. With 'EFFECT OF WAVES = NO' I don't have this problem.

Have you guys experienced this? Could it simply be that the computational time is drastically increased with this process included?

That shouldn't be the case. It doesn't have a significant impact on Sisyphe's runtime. However, with wave effects transport rates are increased dramatically (and hence evolution) so this may create trouble in the hydrodynamics. Enable DEBUG mode and check where the computation 'freezes'.
Costas

Hello,
Do you have the same behavior by using different sediment transport formulas ? (don't forget to use a formula designed to deal with the influence of waves).

pablo

I tried the debug mode and I think your analysis was correct Costas, it "froze" right after updating the 'APPEL DE CORRECTION_DEPTH_3D'.

However, I followed Pablos advice and tried another transport formula. I changed from 4: BIJKER (bed load + suspension) to 8: BAILARD (bed load + suspension) and the computations ran perfectly.

I guess both should work, but for now I guess I'll have to work around the problem by simply using Bailard.

Many thanks for your advice!

schmirre

This workaround should not ease your concerns. Changing transport formula means that the magnitude of sediment transport also changed. If the rate was reduced, the real issue was suppressed for the moment and could emerge later in the computation or if the forcing conditions increase in magnitude. If you have the time, look more into it.
Costas
P.S.
Try the Dibajnia-Watanabe formula that calculates total transport without suspension (if it fits the scope of your study). I have found that bedload calculations are simpler and more robust than bedload+suspension.