GPM VS. PSI FOR A GIVEN PIPE DIAMETER 4

[pencilgeek]

Let me start by saying I don't know anything about fluid mechanics. I need to know how many GPM I can push thru a 2" pipe at 100 PSI. Assuming a length of 10 ft. And what would be the velocity. Seems like it would be a simple calculation. Can someone give me a formula?

Thanks
Chuck

 

[bimr]

pencilgeek

Your question seems to be how much capacity does a hose carry, not a piping question. Ask your rep at Grainger to tell you the answer.

Note that you may need to use multiple hoses as well to obtain the flow rate that you desire.

Note, you will get less flow out of a hose because the hoses may be corragated with poorer hydraulic parameters than pipe.

 

[MikeHalloran]

A fluid mechanics class won't help.
A fluid mechanics degree wouldn't help either; it's not what it sounds like.
This is advanced engineered plumbing.

Buy the Crane book; it's not expensive. Don't lend it out; it has a habit of not coming back.
There's an Imperial version and a Metric version; buy whichever one suits you.

The book will eventually yield to study, and will give you techniques for estimating the pressure drop from end to end of one pipe.

Multiple pipes in parallel or series can be estimated using Kirchoff's Laws, which come from electrical engineering but work just fine for fluids, except that fluid resistors are square-law devices.

If you're using a centrifugal pump, you can't just pick a flow number out of a catalog.
You have to model the entire system for a guesstimated but completely arbitrary flow. Once you have a pressure drop for that arbitrary flow, you can use it to compute what is effectively a Cv, then use the Cv to plot the system resistance curve, and graphically find where the pump curve intersects it, which is the operating point for the pump. You can do a lot of it in Excel. The fancy tools just make a solution quicker.

Luckily for you, the typical mold cooling system is normally bled of air, so you won't have to worry about two-phase flow. Not so lucky for you, the typical mold cooling system has a fairly complex internal geometry. You can model it as a combination of orifices and pipes, but when you test your model against a real mold, it won't be super-accurate. I.e., you'll be doing well to get within a factor of 2 either way. Luckily for you, centrifugal pumps are not super fussy about that.

First buy the book, find the examples that apply to you, and work them.







Mike Halloran
Pembroke Pines, FL, USA

 

[Latexman]

Mike nailed it. You don't need a class. Buy Crane TP410 and study it. It won't be all you need, but it's an excellent start. After you think you know it, go through the examples. It has a chapter of examples worked out for you. It's great! It's concise. It won't take long. A degree in fluid mechanics? I don't think there is such a thing. Some of the best fluid mechanics I know are Chemical, Mechanical, Aerospace, Civil/Environmental and Petroleum Engineers that focused on the area and taught themselves. That's all. You will meet some of them on this forum. When you are ready with specific questions, we'll be here.

On the complex components, like the plastic injection mold, you may need to drill down into the company that furnished it to find the one Engineer that designed/knows it. He/she will probably know how to characterize the resistance to flow, whether it's a K value or Cv value or two K or three K or equivalent length or whatever. If not, you may have to do some flow tests in the field. If so, yeeeehaaaa! You'll really learn something then.

Also, learn to use Search (between Forum and FAQs) to find old threads on the subjects and keywords you need. It takes time, but it's a gold mine!

Good luck,
Latexman

 

[katmar]

The Crane and Cameron books are excellent, but there is no "recipe" in either of them that you can apply mechanistically to solve your problem. Studying either of these sources thoroughly over a few weeks would probably give you most of the required knowledge that you would get in a college course. But you might still be lacking the on-the-job learning and experience that would be required to tackle a real-world problem like this.

One of the biggest conceptual hurdles to overcome it to understand the difference between pressure and pressure drop. Almost all engineers who have worked in fluid mechanics for a few years have this understanding so deep in their psyche that they cannot believe that it is not obvious to everyone. But it is not obvious.

The 100 psi you have is NOT all available to drive the flow through the 10 ft of hose. Part of that 100 psi will be used up overcoming friction in the hose, but some also has to be used to drive the water through your molds and through the return piping. There are techniques in Crane and Cameron for modeling these sequential resistances to flow. A resistance coefficient (usually just called the K value) has to be determined for each section, and for series flow the K values are additive. This is analogous to resistances in an electrical circuit (but more complicated).

If you can measure the pressures at the start and end of your 10 ft of 2" hose then you can calculate the pressure drop across the hose. And once you have the pressure drop you can calculate the flow rate. At 200 gpm the pressure drop would only be 5 or 6 psi across this short piece of hose. Fortunately conservation of mass means that if it is flowing through the hose then the same quantity is flowing through your mold. This means that you can calculate the flow through any section of the circuit for which you have the required information, or as a worst case you have to work with the K value for each element and the overall pressure drop.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"