Avarage Calculation time for centrafugal pump

[B4Tack1000]

Hello

I am running Flowsimulation on a air pump with a twin disc design and outlet pipe. The housing is 500mm the outlet pipe is 1.5 m long.

I started the Calc and it looks like it will take almost 24 hours to complete.

Is the normal? what make it take longer in run time?

I have 5 goals set. Max Vel, Min Vel, Mass flow rate, max total pressure, min total pressure.

The PC is running at 4.1MHz with 12 gig or ram running just under 2000 MHz.

I have run small test with single disc and paddles that take a few minutes, so I'm thinking was up......

Any advice greatly recieved.

 

[kellnerp]

24 hours is not unusual.

2GHz is pretty slow by today's standards.
How many processors are active?

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

http://www.linkedin.com/in/engineeringtransport

Phenom IIx6 1100T = 8GB = FX1400 = XP64SP2 = SW2009SP3
"Node news is good news."

 

[B4Tack1000]

Hello

Ok thanks for that, but I 2000 MHz on the ram is that so slow?

As for the CPU, the PC has 12 and all of them were well used and the ram was maxed out sometimes.

I did get the results, just getting to grip with the data now.

Would you know how to set the arrows on the visualization section to start outside the main chamber and then be dragged into the chamber to represent more real life? Rather than start just on the lip of the chamber

 

[B4Tack1000]

Hello Little video of the result.

Hope it works.

 

[kellnerp]

How fine is the grid? It can be adjusted. Start coarse and refine till the answer converges.

It has been a while since I used FloWorks.

CFD is very compute intensive.

Is this subsonic?

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

http://www.linkedin.com/in/engineeringtransport

Phenom IIx6 1100T = 8GB = FX1400 = XP64SP2 = SW2009SP3
"Node news is good news."

 

[B4Tack1000]

Hello

Brief update, the reason it seems for the long calculation was a small misalignment causing a ridge. Once removed the calculation is about 30 minutes now.

With mods I have found the areas for improvement and have gained 16 M/S over the base number. So, very pleased with that.


The grid was default set to 3. I have run over 150 tests on this design, early ones to learn the program and the later ones to get the increase in performance.

I have any project to solve and have need for some advice on that one too.

I have been asked to evaluate a Plenum chamber on a turbo charged Race Bike. I have seen on YouTube a way to simulate the 4 outlet ports opening. I have completed the solve section with the vector arrows changing direction in time with the ports opening. However when I run the video it only plays the last section IE only one port open. I can't seem to get the whole test run in animation.

Any ideas?
Thanks.

 

[B4Tack1000]

Hello I am think it best to start a new one on the Bike Plenum as to not confuse this post title.

Go here:

http://www.eng-tips.com/submitpost.cfm?pid=1183

 

[B4Tack1000]

Hello

small clip of the final test, I think you can see for yourself the improvement from the first video.

http://www.youtube.com/watch?v=TbujHACDCSQ&feature=youtu.be

 

[kellnerp]

it seems for the long calculation was a small misalignment causing a ridge.

This is why using TOOLS/CHECK with short edges and gaps is a good idea.

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

http://www.linkedin.com/in/engineeringtransport

Phenom IIx6 1100T = 8GB = FX1400 = XP64SP2 = SW2009SP3
"Node news is good news."