Mixing water in ANSYS

I will be using this study case as a tool to learn the side by side comparison that ANSYS allows me to make. This is a powerful tool for someone who wants to develop a component or try and optimize the performance of one. Many times I would think to myself “what if I changed this here” or “what if that other dimension was different”. CAD in general and CAE especially makes this process a lot simpler than I initially thought. This process is cumbersome when performed in SolidWorks (at least the 2016 version that I have access to), but ANSYS offers a quick result rendition and point probing tools.

First, let me go over the project setup:

Figure 1 – ANSYS Workbench setup

To those familiar with ANSYS, you would usually see just one column per simulation. The normal flow exists. The [geometry -> mesh -> setup -> solution -> results] has been calculated for each of the two geometries. To make everything easier I did one complete setup for a problem, then duplicated the project and changed the geometry file, then generated the mesh again, ran the calculation, and it was done. Since all the options were already set it saved me the time it would take to set up a whole new problem (which is usually a lot). On top of that, I could repeat this for numerous different changes so I could decide which one is best for the problem and what we’re trying to achieve.

As good experiments go, when comparing results everything should be equal except for one variable, in this case, it’s the hot water inlet diameter. The small one is 25.4 [mm] {or 1 inch in imperial). The second version is 38.1 [mm] in diameter (or 1.5 inches in imperial).

Let me explain the problem itself first. As we see in figure 2:

Let’s call the one on the left Elbow2 and the one on the right Elbow

Figure 2 – Geometry

We have the main pipe, connected to it is an auxiliary pipe. The cold water inlet is on the left, the cold water enters at 293[K], in the direction of +x (red). The hot water inlet is at the bottom of the geometry, hot water enters at 313 [K] in the direction of +y (green).
The cold water enters at 0.4 [m/s] normal to the face of course.
The hot water enters at 1.2 [m/s] normal to the bottom pipe for Elbow, and 0.8 [m/s] for Elbow2.

As I was writing this I was wondering “Should I leave the mass flow rate equal or should I leave the inlet velocity equal?” The flow rate is, of course, the cross section area X normal velocity. The result is that there’s 50% more mass flux of hot water into the cold water in comparison to the smaller pipe.

The answer to this was not simple, I had to change the boundary conditions for Elbow2 so that the flow rate is matching. I’m coming up with a few different reasons for why each of these differences might be the correct one. But I’m going with equal mass flow rate.


Let me just quickly go over the mesh, I won’t in depth into the mesh, as we remember, I want the two cases to be as similar as possible, which means the mesh needs to be equally simple for both.


Figure 3 – Pipe connection mesh



Figure 4 – Hot water inlet


The whole geometry is divided into tetrahedrons with one exception, you can see all the boundaries are lined with a certain directional bias, this was dictated by what you can see in figure 5. This feature is there to take care of any boundary flow and the no-slip condition, then as we get closer to the middle of the pipe it is less significant to have a Hexa mesh.

Figure 5 – How to create the layers near the boundary

One last thing, before the results, often when performing analyses of this type we see that only half or a quarter of the geometry is taken into the analysis. In this analysis, for example, only half of the geometry has been modeled, as is shown in figure 4. This aids us to greatly reduce calculation times, much like when programming an integral calculation. When we have (O^n) which represents run-time complexity, and dictates how long it takes a computer to calculate



As always, I go straight for the prize, this time the temperature distribution

side by side temp1
Figure 6 – Temperature profile

Temperature distribution along the different pipes, blue is about 292 [K] and red is about 313 [K]. mixing is obviously better on Eblow rather than Elbow2



side by side velocity
Figure 7 – Velocity profile





Pressure side side
Figure 8 – Relative pressure



Final thoughts

This example is a visually pleasing to me, it’s a nice representation of what’s going on when mixing two fluids that have different temperatures (which also entails different viscosity and density) how they interact and mix to become one uniform liquid.

One thing that immediately caught my eye was the fact that in the small hot inlet (Elbow) as opposed to the larger one (Elbow2) the water do not mix well in the latter. The flow remains laminar and separated, it WILL mix later on. In a case where I would have limited space for mixing I would think about creating some turbulence in order to mix the fluid better or maybe creating a longer pipe path or change the diameter along the way to ensure the fluids mix well.

This could represent for example a faucet at home (well, a weird one but bear with me), which works on the principle of mixing different volumes of room temperature water and hot (varies from home to home) water around 70 degrees Celsius. but we will continue later about how taps and faucets work.



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s