[Shamoon Jamshed]

Dear Fellows,

For periodic you go to mesh/modify-zones/make-periodic. You need to have a rotational axis and two planes for rotational periodicity(and they should be similar in rotation), but in my case the left and right planes are not similar, in fact the right one is a bit translated in z (z is rot. axis) and start after some distance along z from the left plane, now if the left plane is rotated it would never be the same plane as the right one. I have now tried to model it as interface and then specify ROTATIONAL PERIODCI. But I dont know abt it much. Please let me know if any one has experience of that.

[LuckyTran]

Why are you using a rotationally periodic b.c.? It seems you should be using the simpler translational periodic b.c.

[Shamoon Jamshed]

Originally Posted by LuckyTran

Why are you using a rotationally periodic b.c.? It seems you should be using the simpler translational periodic b.c.

Dear Lucky,

let me elaborate my case a bit more. The thing is , that I have a helical groove that is etched on the surface of a tube. Now imagine a number of grooves starting from one end of the tube. So these helix are grooved till the end.
In one paper I read that for computational analysis, one may take rotational periodicity. Now in fluent, in order to specify the rotational periodicity, the periodic ends should have to be overlapped. This is not in my case, unfortunately. The reason is that I have a helical geometry and helix has a pitch angle distance. So the two periodic planes (Lft and Rt) can never overlap due to a segment of helical groove starting from left and going right. Second, after moving along the length (about a pitch length) we will meet another groove that comes from the second helix. Imagine the micrometer screw gauge and its ratchet. The segment I took is 36 degree segment out of 360 because there are 12 grooves starting. Please see the attached file

[LuckyTran]

Originally Posted by Shamoon Jamshed

Dear Lucky,

let me elaborate my case a bit more. The thing is , that I have a helical groove that is etched on the surface of a tube. Now imagine a number of grooves starting from one end of the tube. So these helix are grooved till the end.
In one paper I read that for computational analysis, one may take rotational periodicity. Now in fluent, in order to specify the rotational periodicity, the periodic ends should have to be overlapped. This is not in my case, unfortunately.

Ok. I thought you were trying to do the periodic b.c. in the streamwise direction. Now I get what you are trying to do.

Correct the helix must overlap. The trick is to do it as shown in the picture that you attached. The left and right planes are not flat, they are twisted to match the twist of the helix. If you twist your planes as shown then you will be able to match the left and right planes with only one helical groove.

It is also possible to do it with flat planes but you need to make the domain long enough in the streamwise direction so that the helix is repeated. Doing it this way there will be multiple grooves in your domain. Just play with around a bit, you should get it eventually.

[Shamoon Jamshed]

Originally Posted by LuckyTran

Ok. I thought you were trying to do the periodic b.c. in the streamwise direction. Now I get what you are trying to do.

Correct the helix must overlap. The trick is to do it as shown in the picture that you attached. The left and right planes are not flat, they are twisted to match the twist of the helix. If you twist your planes as shown then you will be able to match the left and right planes with only one helical groove.

It is also possible to do it with flat planes but you need to make the domain long enough in the streamwise direction so that the helix is repeated. Doing it this way there will be multiple grooves in your domain. Just play with around a bit, you should get it eventually.

Thanks Lucky,
Now the problem is that Fluent only rotates about the conventional axes (x,y or z) and not the user specified axis. If that was the case, (fluent allow me to specify my axis) my life would have been much easier. Now I have z axis as rotation. And in axial direction there will be around 500 grooves because pitch angle is very large.
Please explain your sentence
"It is also possible to do it with flat planes but you need to make the domain long enough in the streamwise direction so that the helix is repeated"
thanks a lot for answering!

[LuckyTran]

You can specify any rotation axis, it's in the cell zone conditions. Fluent computes the rotation angle based on this rotation axis. If you specify the incorrect axis, then you'll get an error "Failed to make zones periodic."

I'm not sure where you are getting 500, there are only 12 grooves. No matter how you slice it, there can be no more than 12 grooves.

Using two slices, you can generate a periodic domain. Though you'll slice through more than one groove. The maximum number of pitches you'll need to be able to generate a domain is the number of grooves, 12. But you may be able to exploit other symmetries and have less grooves.

[Shamoon Jamshed]

Originally Posted by LuckyTran

You can specify any rotation axis, it's in the cell zone conditions. Fluent computes the rotation angle based on this rotation axis. If you specify the incorrect axis, then you'll get an error "Failed to make zones periodic."

I'm not sure where you are getting 500, there are only 12 grooves. No matter how you slice it, there can be no more than 12 grooves.

Using two slices, you can generate a periodic domain. Though you'll slice through more than one groove. The maximum number of pitches you'll need to be able to generate a domain is the number of grooves, 12. But you may be able to exploit other symmetries and have less grooves.

Using default axis or (customized axis) in fluent will only be used if we make periodic bc through TUI.
Here what I asked initially as well, that I made interface on the two planes and specify them as periodic boundaries.
by 500 or so, I meant that if I take slice, there would be first groove then after some distance the groove from the second one will come then the third and so on. By the time all 12 grooves will come, the pitch of the first groove will complete, lets say it is 15mm. My desired pipe length is 215 mm and so you can do the math. 15mm has 12 grooves so 215 mm will have 15*215/12 ~ around 260 grooves. Ok, I roughly said 500 but here it is 260 grooves, this is correct value now.

See the attached grid, I made upto 20 mm.