My company makes a water filled material handler that consists of a reservoir load tank and a channel conveyor. There is a pump to regulate the conveyor flow and recirculate the water back to the loading tank. Bins of material are forced underwater, the material floats out and is carried by the current. A recent change in the design has caused surge waves in the conveyor channel when bins are put in the tank. We have moderated these waves by a trial-and-error process, but the water still surges.

What formulas, software, books, etc., can anyone recommend to model the waves or water surface of the channel flow under these conditions? I'd like to minimize the waves still further, and also gain design experience so this doesn't happen again. Specialized software like Fluent or Flowtran is out of the question. Hard work is OK, I just am not sure where to start.

The model could be 2-dimensional as a section of the rectangular channel, and believe the the underlying flows and turbulence are not a great concern, as the material floats in the water. Although it's a transient wave, the floating material damps it so that the third wave (second "bounce") is negligible.

(1). Dr. Tony Hirt is the world leading expert in free surface CFD problems. (2). I think, you can visit his company's website (listed as sponsor in this forum), Flow Science.Inc.

Thanks for the response. Perhaps I didn't make myself clear that this type of dedicated, expensive software is not going to be purchased. In the case I cited, it's cheaper to send a mechanic to the site with a welder and some scrap metal for few days to "try some things". As it happened, it worked -- pretty much. I'd prefer to get a good estimate of what will happen before we make it - if we make another similar system.

Perhaps I'm off base, but the modeling of a surge wave (and damping effects) doesn't appear to be overly complex -- within range of the average calculus-able person. My knowlege is mechanical, and I'd appreciate pointers to fluids modeling.

(1). Dear friend, please don't jump to the conclusion. (2). Do you think that I am sending you a bill based on the time spent? (2). I don't know whether Dr. Hirt will have the time to answer your question (I mean, give you some design hints)? If he does, then his answer will be more reliable than anyone else. (3). We almost never answer a question based on the results of a computer run. It is based on the commonsense and experience. Why don't you send e-mail to Dr. Hirt, it is quite possible that his few words of suggestion will keep your mechanic very happy. (4). I can try to give you some suggestions, but you will have to outline your systems and problems more clearly. Anyone with similar system experience is encouraged to provide some help.

(1). AS the bins are lowered into the tank, it will create waves. Then it will follow the stream and move into the duct. On the surface, there will be some oranges and apples riding the waves and floating into the duct. The question is how to reduce the wave. (2). The solution-1: gradually lower the bins into the tank to release the oranges and apples. (3). The solution-2: increase the size of the tank, such that the location between the bins and the entrance to the duct is longer. (4). The solution-3: install a wire cage between the bin location and the entrance of the duct as the wave breaker. The wave generated will have to go through the wave breaker to reach the entrance of the duct. The round cage will guide the oranges and apples around it and enter the duct from both side. (5). The solution-4: design a guide along the tank wall to the entrance of the duct, such that oranges and apples will travel along the wall inside the guide first before entering the duct. The circular section of the guide (assuming the tank is cylindrical in shape) will provide more wall resistance to damp the wave in addition to the wave reflection between the circular tank wall and the circular guide wall. (6). That should keep your mechanic busy for a few hours. I have to go now. good night.

If hard work doesn't bother you, and you have some time and inclination to spend learning about fluid surface modelling, then there are some codes out there that will be able to help. As John mentioned, Dr. Hirt is one of the foremost leaders in free surface modelling. I believe you can still get the source of one of the original codes that Dr. Hirt developed while he was at Los Alamos National Lab. Go to the online library at the LANL web site. The base code is called SOLA and there exists a free surface extension to it called SOLA-SURF. I have worked with both SOLA and SOLA-SURF in advanced CFD courses and have done modelling similar to what you are looking for. Be aware, however, that SOLA is a rudimentary CFD code with no fancy features, and there will be some customization needed to properly pose your problem. If this is a business problem where you need an answer in a timely fashion, then I would explore the possibility of contracting Flow Science. If this is a longer term, interest only project, then good luck.

Hi There,

I am not familiar with the floating apples and oranges in the water tank, etc... but if I had to model such a thing I would do some 2D Shallow Water Approximation to model the surface waves (inertio-gravity waves). You can find some info on that on books in Geohysics or even just Hydro, etc..

Patrick G.

The problem was a business problem, but was alleviated by "hacking" on it until it was acceptably small. I wanted to understand the process so that I could design a method of solving (dissipation, negation) the problem if and when I made another fluid conveyor.

Los Alamos seems to feel that SOLA is a state secret, but I will continue to search the site until I am arrested <grin>. Perhaps naively I thought that there would be some equations that could be used for this problem. Is "code" -- as in the SOLA code -- the raw equation(s) or uncompiled program code?