[lakku12]

I am studying the peristalsis on the ureter. As we all know the ureter is not oriented with respect to one axis. I want to simulate the bolus (peristalsis) motion on the wall.
I need some help or suggestion
How it to be simulated?

[ghorrocks]

You obviously need to define the motion, as this is fundamental to this flow. If you cannot define the motion you cannot simulate it.

The two main approaches are:
1) To define the motion directly and use moving mesh. You will need to know the motion of the walls over time (sounds tricky), and you won't get any coupling of the motion from the fluid flow (which is probably significant in this case). But it is the simplest approach if you can do it.

2) Use FSI (fluid structure interaction) to move the mesh. This means you can generate the motion by doing a physically realistic action (is it a muscular motion? Then you could put a muscle type element around it and contract it). The motion is then the result of the action, and includes coupling with the fluid flow (so is more accurate in some cases) but will require you to couple CFX to ANSYS Mechanical using FSI which is a more complex simulation.

In your case I suspect option 2 is required. This means you will have to specify the muscle motion in ANSYS Mechanical somehow.

[lakku12]

Thank you Glenn Horrocks,

I done the simulation for straight tube. But I want to simulate it for the axis which in not oriented with respect to one definite axis. I don't know how to simulate or apply the load which run through the axis.

https://drive.google.com/file/d/1yTY...ew?usp=sharing

[ghorrocks]

The concept of the curved path is the same as the straight path, you "just" need to apply a transformation to the curved path. I say "just" as the transformation is not necessarily simple.

But it is a simple choice - if you cannot define the transformation then you cannot model it.

Then you might want to consider the option #2 I mentioned. If you do it as an FSI model then you "just" have to get some elements around the tube to squeeze in a rippling fashion. This approach probably will not require a complex geometric transformation function, but it introduces a whole bunch of other complexities instead.

And finally: does the curved path make any difference anyway? If the flow is a low Reynolds number flow then the only important factor is the cross section and length. The curve might cause an insignificant difference.