Hello all,

the result of a cfd job of the flow in a 90 degrees bend with a rectangle cross section showed at 45 degrees a higher velocity at the inside and a higher velocity at the outside. For my understanding this result is qualitatively right. But I have got some doubts after all collegues in my department tell me that in reality the highest velocity is at the outside and the lowest at the inside even at 45 degrees. But I think due to the positive pressure gradient the fluid at the outside has to be decelerated till reaching the pressure maximum. The same rules for the situation at the inside the other way round. Am I wrong with that thoughts ?

Thanks,

Tim

Little mistake in my first question: The CFD results showed a lower velocity at the outside and a higher at the inside of the bend.

Sorry,

Tim

(1). First, you have flow moving forward in a rectangular pipe. (2). The pipe then makes a right 90 degree turn. (3). The fluid particle on the left side of the pipe will find the pipe wall in front of it. (4). The fluid particle on the right side of the pipe will find the space is opening up quickly. (5). The fluid particle on the left side will have to slow down and make a gradual right turn, otherwise, it will hit the wall or go through the wall (if the wall is made of paper material). (6). The fluid particle on the right side of the pipe, finding extra space suddenly, tends to fill the extra empty space and make a quick right turn. This action will lower the pressure in low speed flow. (7). When the left side fluid particle slows down, the pressure will increase, similar to the stagnation point region.

With the danger of offeding your collegaues in the department, I should say you're right. For the flow near the outer wall at 45 deg, it should be an uphill interms of pressure gradient, leading to a deceleration. For 3D, the situation can be a little more complicated and the maximum velocity can shift toward either the inner or outer wall depending on where you're depthwise.

John,

what would be your guess of the flow conditions at the end of the bend. If no separation at the inside occurs, will the velocity profile be uniform again because the outer fluid is accelerated and the inner fluid is decellerated by the pressure gradient in flow direction ??

Tim

(1). Now you are talking about the detailed flow development in the pipe. (2). I think, it will depends on many factors like, the initial flow non-uniformity, the Reynolds number, the shape of the bend, the radius of the bend, the aspect ratio of the rectangular pipe,etc.... You definitely will have a long separation bubble, if it is a 90 degree sharp bend. (3). You can look into these interesting results by running 3-D CFD calculations. (if you have the time to do so.) (4). In 3-D bend, you will have secondary flow motion also.

Tim,

I would also agree with your results rather than with your colleagues' opinion. I did some investigations on a 180 degree bend and the "movement" of higher velocities towards the outer bend began definitely after 45 degrees. sometimes, depending on the BCs, even after 90 degrees. hope this helps

thomi