Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Laminar flow model with two inlets, one outlet, and a vacuum pump

Please login with a confirmed email address before reporting spam

Hello. I am new to comsol.

I am currently trying to model a cylindrical region in COMSOL (Laminar Flow Interface) that operates in the laminar flow regime. I am using one side of the cylinder to input Helium gas while simultaneously pumping on the other end. Half way down the cylinder is a pinhole that acts as an outlet. There is also a water drop in front of the pinhole which is intended to model a water droplet (the sample). I have set this drop as another inlet to simulate the evaporation rate. Finally, I set the outlet pressure to about 1 Pascal to simulate a region further down in the system (not part of the geometry).

The goal of this simulation is to set the evaporation rate of the water as constant and determine the most efficient input rate for the helium (a.k.a. the lowest) and pumping rate for the vacuum while maintaining a pressure of about 10 mbar (1000 Pa) throughout the cylinder (except at the pinhole). Once those rates are determined, I need to measure the flow rate out the pinhole.

I have been able to set the model up so far but there are several things that I am unsure of:

What exactly is the [static pressure at no flow parameter] for the vacuum pump?
My understanding was that it was the pressure associated with the throughput (q=P*dV/dt) of the vacuum pump. But this value seems to set the pressure of the entire cylindrical region in my simulation. This could be as expected but I wasn't sure if I was interpreting that parameter incorrectly.

I have currently set the wall boundary conditions to no slip which was the default but I wasn't sure when the slip condition was applicable. What are some example situations where the slip condition should used?

I have set the outlet pressure to the pressure that is expected in the next region (~1 Pa), but I have noticed that the pressure on that boundary never reaches that value once I run the simulation. I initially thought that this meant that the outlet pressure I set was too low and that the pressure shown was the lowest pressure attainable through that pinhole. But then when I kept all other parameters constant and lowered the outlet pressure again, the simulation computed a lower pressure on that same boundary but still not equal to the outlet pressure I set. Does anybody understand what might be going on here? Is COMSOL able to adjust boundary conditions?

My biggest issue is how to interpret the change in pressure when I change certain parameters. The pressure tends to move in the opposite direction than I would have thought. For instance, when I increase the flow through the top inlet, the average pressure seems to go down. And when I increase the vacuum pump flow rate, the average pressure goes up. I may be misunderstanding something (and likely have something set up incorrectly) but isn't this counter intuitive? Shouldn't the pressure increase if the flow in is higher and the flow out is lower?

Lastly, I am currently making the assumption that the entire domain is composed of Helium gas for simplicity but in reality the gas emitted from the droplet is water while the other inlet is inputting Helium. Is it possible to make the region a combination of water vapor and Helium gas with different concentrations? Or possibly would it be more accurate to simulate the drop as actual liquid water instead of spherical inlet?

Help with any of these questions would be much appreciated.

7 Replies Last Post 24 feb 2017, 12:05 GMT-5

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 21 feb 2017, 14:06 GMT-5
I forgot to add some pictures so I have attached them to this post.
I forgot to add some pictures so I have attached them to this post.


Please login with a confirmed email address before reporting spam

Posted: 8 years ago 22 feb 2017, 04:18 GMT-5
Hello

I am not sure if I understood fully what you want to simulate but I think you are missing a few physics interfaces for what you want to simulate.

If you want to simulate evaporation you might need the transport of diluted species physics. Here is an example:

www.comsol.com/model/evaporative-cooling-of-water-6192

However, I think as you use a vacuum pump you are sucking helium into your tube. This should mean that the water will also be sucked in through the inlet and create droplets in your gas flow
If you want to simulate this phenomenon you need the two-phase flow interface:

www.comsol.com/model/capillary-filling-phase-field-method-1878

You should use the no-slip condition. Sliding could happen if your gas pressure is really low (which is not the case here)….

All vacuum pumps have a pressure vs. flow curve (given by the manufacturer). As you start pumping there is a lot of gas present in the chamber and there is nearly no pressure difference between the two sides of the pump, so gas can flow easily (free delivery flow rate). As you continue pumping the pressure difference becomes bigger and the flow rate decreases until the pump is incapable of pumping any more gas out (flow = 0). This pressure is the maximum pressure difference achievable with that certain pump (static pressure at no flow)

I hope this answers some of your questions
Hello I am not sure if I understood fully what you want to simulate but I think you are missing a few physics interfaces for what you want to simulate. If you want to simulate evaporation you might need the transport of diluted species physics. Here is an example: https://www.comsol.com/model/evaporative-cooling-of-water-6192 However, I think as you use a vacuum pump you are sucking helium into your tube. This should mean that the water will also be sucked in through the inlet and create droplets in your gas flow If you want to simulate this phenomenon you need the two-phase flow interface: https://www.comsol.com/model/capillary-filling-phase-field-method-1878 You should use the no-slip condition. Sliding could happen if your gas pressure is really low (which is not the case here)…. All vacuum pumps have a pressure vs. flow curve (given by the manufacturer). As you start pumping there is a lot of gas present in the chamber and there is nearly no pressure difference between the two sides of the pump, so gas can flow easily (free delivery flow rate). As you continue pumping the pressure difference becomes bigger and the flow rate decreases until the pump is incapable of pumping any more gas out (flow = 0). This pressure is the maximum pressure difference achievable with that certain pump (static pressure at no flow) I hope this answers some of your questions

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 22 feb 2017, 12:39 GMT-5
Thank you for responding. This was very helpful.

For the issue of the drop being sucked in through the vacuum pump, this will likely not happen because the drop will been fixed more or less in place with a feedthrough device. Also, the pumping speed will be maintained at relatively low speed. The main purpose for pumping on the system while simultaneously feeding Helium through the inlet is to essentially replace the air molecules inside the system with Helium molecules. The Helium molecules are more beneficial for an application of the system because of their smaller molecular diameter.

And for the evaporation rate, I don't really need to simulate it because I have already measured that experimentally. The drop is mainly included to observe the affect it has on the outflow through the small pinhole since the drop is fixed directly in front of the pinhole.

I do believe there is still something wrong with my settings since the trend I am observing does not make any sense. When I increase the flow in through the inlet, the pressure goes down. And when I increase the free delivery flow rate, the pressure goes up.
(I am treating the free delivery flow rate as a variable in order to determine which pump I will need to order)
Thank you for responding. This was very helpful. For the issue of the drop being sucked in through the vacuum pump, this will likely not happen because the drop will been fixed more or less in place with a feedthrough device. Also, the pumping speed will be maintained at relatively low speed. The main purpose for pumping on the system while simultaneously feeding Helium through the inlet is to essentially replace the air molecules inside the system with Helium molecules. The Helium molecules are more beneficial for an application of the system because of their smaller molecular diameter. And for the evaporation rate, I don't really need to simulate it because I have already measured that experimentally. The drop is mainly included to observe the affect it has on the outflow through the small pinhole since the drop is fixed directly in front of the pinhole. I do believe there is still something wrong with my settings since the trend I am observing does not make any sense. When I increase the flow in through the inlet, the pressure goes down. And when I increase the free delivery flow rate, the pressure goes up. (I am treating the free delivery flow rate as a variable in order to determine which pump I will need to order)


Please login with a confirmed email address before reporting spam

Posted: 8 years ago 23 feb 2017, 09:21 GMT-5
so are you using the droplet to clog the pinhole?

In that case, I really think you cannot have two liquids with laminar flow physics alone. you have to add two-phase flow (level set or phase filed) to track their interface.
so are you using the droplet to clog the pinhole? In that case, I really think you cannot have two liquids with laminar flow physics alone. you have to add two-phase flow (level set or phase filed) to track their interface.

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 23 feb 2017, 12:29 GMT-5
Essentially, the drop will be hanging from a syringe type of apparatus and particles will be sent through the pinhole and get implanted in the water drop. But I do not need to consider the particles currently. I am just trying to use the rates of flow in and out of the system to determine what the environment (pressure, speed distribution) is like through out the cylindrical region and pinhole.
The drop is mainly acting as another boundary and an input of more molecules.

Are you able to use the two phase flow if the liquid (a.k.a. the drop in my case) is not moving?
If so, then two phase flow would likely be the better option for me.
Essentially, the drop will be hanging from a syringe type of apparatus and particles will be sent through the pinhole and get implanted in the water drop. But I do not need to consider the particles currently. I am just trying to use the rates of flow in and out of the system to determine what the environment (pressure, speed distribution) is like through out the cylindrical region and pinhole. The drop is mainly acting as another boundary and an input of more molecules. Are you able to use the two phase flow if the liquid (a.k.a. the drop in my case) is not moving? If so, then two phase flow would likely be the better option for me.

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 24 feb 2017, 08:41 GMT-5
the drop will remain stationary its surface contact energy is sufficient (i.e. the contact angle that you define in the simulation using the two-phase flow interface. then you can simulate if it stays in place or flies away do to the flow.

however, I think you can estimate the pressures exerted on the droplet using simple formulas for drag coefficinet. so if you drop the water part in your simulation you could extract the values of speed from the simulation to calculate the drag force.

alternatively, you could use FSI (fluid-structure interaction) module to calculate the pressure exerted on a little "solid" pretrusion.

the two-phase flow is the most realistic option but it will need more tweaking to make sure it is giving the correct values (you should have a careful look at the documentation of two-phase flow)

the drop will remain stationary its surface contact energy is sufficient (i.e. the contact angle that you define in the simulation using the two-phase flow interface. then you can simulate if it stays in place or flies away do to the flow. however, I think you can estimate the pressures exerted on the droplet using simple formulas for drag coefficinet. so if you drop the water part in your simulation you could extract the values of speed from the simulation to calculate the drag force. alternatively, you could use FSI (fluid-structure interaction) module to calculate the pressure exerted on a little "solid" pretrusion. the two-phase flow is the most realistic option but it will need more tweaking to make sure it is giving the correct values (you should have a careful look at the documentation of two-phase flow)

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 24 feb 2017, 12:05 GMT-5
Okay. Thank you very much.
I am going to try the two phase flow and I will get back to you if I run in to any serious issues.
Okay. Thank you very much. I am going to try the two phase flow and I will get back to you if I run in to any serious issues.

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.