Hello,

I'm not very knowledgable about CFD. I'm beginning work on a project where I will need to calculate steady state and transient fluid flow in a system that includes compliant piping elements (plastic tubing) that is known to have a significant impact on transient flow. Is this requirement within the capabilities of most commercially available CFD packages? Which packages might be suitable for this analysis? Is there any recommended shareware that might be suitable? Thanks in advance for any help offered.

Ken Wiljanen

(1). What is your need? What are you trying to do? What have you got to offer? (2). If you don't know CFD, then you may start reading CFD books. (3). If you know CFD, but you don't have time to write the code, then you can define your problem and start looking for consultants. (4). If you like to do it by yourself, using a commercial code, then you can visit the websites of the commercial code vendors first and contact their engineers. (5). If you don't have money and time, then it will be difficult to get good solutions. (6). Internet is sort of free, so, you can do the surfing first.

Ken,

You need a code that will perform coupled Fluid-Structure Interaction (FSI). Our CFD-ACE+ code has a stress solver that is fully coupled with the flow solver. We have used this code internally to do a wide variety of coupled FSI simulations, and we have several external users that are using it for simulations involving compliant tubing (blood vessels for the most part).

Please take a look at our web site <www.cfdrc.com>, and feel free to contact me with any additional questions.

Regards, Alton

John,

Thanks for your response. I'm working with a client that manufactures a product for the medical field. This product is one part of a system that is also comprised of a pump (peristaltic or vacuum) and a valve and plastic tubing. I'm interested in the modeling the transient response of the system to being occluded (flow is blocked) and then being cleared of whatever material was blocking flow. Field experience indicates that the compliance of the plastic tubing has a significant impact on transient response. This is a relatively trivial problem, with the exception of the compliant plastic tubing. I certainly have no interest in writing my own code. I am perfectly capable of handling the project with the appropriate software package. My goal in posting my question was to short circuit hours and hours of calling people and sending e-mail in an effort to identify which codes were appropriate to address this problem, since there appears to many different packages available.

Regards,

Ken Wiljanen

(1). It is a very common device or pump, so, there must be somethings done about it already. (2). If you can model the heart and valve, you can also model the pump with moving wall. (3). It is still difficult for commercial cfd codes to have this capability, so check around first.

I'm not interested in modeling the pump. What I'm interested in is the transient response of the system as flow is blocked and then resumes. The pump merely provides suction with a defined performance curve. The plastic tubing will compress when flow is blocked and then expand when flow resumes, causing a surge in flow. I'm learning that this is a difficult phenomena to model.

I'm not interested in details of flow at a very small scale, but rather flow rate at a point or points in the system. In my ignorance it seems to me that what I am describing is analagous (sp?) to a highly damped spring in a mechanical system. I'm afraid that I'm going to have to dust off the books and delve into the math involved. Basically what the tubing is is a variable volume element where the volume is a function of the pressure differential between the tubing interior and the atmosphere, with a spring rate and some damping.

Thanks for your help.

Ken Wiljanen

(1). If you take a flat roller and press down onto the flexible plastic tubing with liquid in it, you can separate the liquid into two parts, one ahead of the roller and the other behind the roller. (2). as you move the roller forward, you will be compressing the liquid forward. (or pushing the liquid forward as a pump) (3). the mass flow rate will be related to the speed of the roller and the volume covered or travelled. (4). Since the liquid is not compressible, it has to go forward. (5). If the outlet of the pump is restricted or is connected to a high resistence system, then there will be liquid accumulation in the flexible tubing portion acting like a ballon. (assuming that the tubing can expand) like filling the water into a ballon. (6). So, even if the liquid is incompressible, like the traffic flow, it is a transient process. In this case, you have (mass flow in - mass flow out)= time rate of liquid accumulation in the expanding tubing section. (7). the amount of mass flow out is determined by the downstream system side, and the mass flow in is given by the roller speed and is known. So, if you stop the outflow, the amount of mass accumulation rate will be equal to the mass flow in created by the roller motion. (8). So, this is a simple transient mass conservation equation. Since the mass flow in is given by the known roller speed, the system is determined by the downstream system mass flow out. This can be modelled as the mass flow rate vs the pressure difference between the tubing side and the downstream system outlet side and the resistence of the downstream. (9). If the flexible, expanding tubing is weak, then it will not create high pressure to push the liquid forward. So, you will have to know the property of the tubing. say volume of the tubing vs the pressure force created. like pulling a spring or a rubber band. Once this is known, you have the equation between the amount of liquid in the tubing vs the pressure created. The larger the liquid volume in the tubing, the higher the pressure created. (10). If you modelled it linearly, it becomes P,1 = const * Volume_of_liquid_accumulation_rate. From P,1 and P,2 you can determine the mass flow rate out through the downstream system using the system resistence formula, which must be given also. (11). If you get these variable together, you can find the transient solution. Hope you can see my point. (it is like the traffic flow simulation, except that this time, the spring force in the tubing section also affect the outflow rate.

> Basically what the tubing is is a variable volume element where the volume is a function of the pressure differential between the tubing interior and the atmosphere, with a spring rate and some damping.

I believe you are on the right track here. Based on the description of what it is that you are looking for, you are basically trying to solve a transient 1-D problem. So you need to take a time-varying control volume and make sure that you apply the correct boundary condition. BTW, the boundary condition is perhaps the most difficult part of your analysis because you need to know some things about the tubing. But you would need that for CFD as well anyway. So, you'll probably end up with a relatively simple system of equations that you can solve using a few lines of fortran or other math packages. As a starting point, you want to check out Schlichting; there is a little section on modeling pulsatile flow (not exactly what you're after though). You probably can find something similar in a relevant textbook.

Adrin Gharakhani

Thanks for all of your responses. Just as I was getting into this problem, I've been reassigned to another task for 4 to 6 weeks. I may pick up this thread again at that time. Thanks for your help.

Ken Wiljanen

Ken,

I second Alton. Try to get some more information on CFD-ACE+. I am sure that you can get plenty of examples from them that can demostrate the capability of the code to do couple FSI.

raza.