Thread: Ultimate DIY Separator / Extractor

Originally Posted by HeadScratcher

Certain parts of the cyclone design can’t be changed due to that being the way it works, so we are fairly well locked on the flow restrictions caused by the inner and outer tubes (but not size), the only variables we have to play with are the collection areas before and after the cyclones…
If you look at the last picture I posted, the 6 tubes dump directly into the 150mm pipe (no step down now), so really the clunky transition of multiple pipes into 1 has been removed, as long as the flow of the pipes feeding the 150mm pipe exceed the flow of the 150mm pipe there should be minimal flow impact.

Unless there is a graded transition of a smooth wall from one pipe to another over a pipe length of 5 - 7 pipe diameters of the bigger pipe the impact will be significant. Abrupt transitions, especially one that just involves inserting one pipe into another are amongst the worst ways to construct a transition.

To improve flow into the 150mm pipe I could trim the joining pipes flush to the inside of the 150mm pipe (once glued and set)… Can you run the math on 5 x 86mm pipes to see if that would be adequate flow for 1 x 150mm (146mm ID?) outlet pipe please.[/FONT][/SIZE]

Yep 5 x 86mm pipes will do it but like I said this is likely to be the least of your worries.

Ok so perhaps instead of two sharp 90 degree turns, how about two 45 degree bends and come in from the bottom that way the flow is going with the 150mm pipe. Would make the lid tricker but gives much better flow.

However without doing the math I am thinking you wouldn't get 5 x 90mm holes into 150mm hole, so perhaps coming into the side of the 150mm pipe at 45 degree angles. Still cuts out 90 degrees bend compared to two 90s.

Edit:
or maybe even this might work...

2 x 45º's are a bit better than 1 x 90º but the biggest problem in those designs is the dead end block and small pipes coming out of the sides of the bigger pipe. I have highlighted were the problems are in red and suggested a more efficient type if transition profile in blue on the LHS diagram. This would have to be designed aerodynamically and made quite precisely or it could lead to even more turbulence.

I thought I had posted a picture showing an upside down cone in the centre, but can't seem to see one in any of the pics.

Be quite a bit trickier, but if I had all 5 x 45 degree angles coming throught the bottom / side so they just touched, with a cone sitting in the dead spot in the centre it would probably be as optimal as it could be made with hand tools (short of custom molded plasted)

Originally Posted by HeadScratcher

I thought I had posted a picture showing an upside down cone in the centre, but can't seem to see one in any of the pics.

Be quite a bit trickier, but if I had all 5 x 45 degree angles coming throught the bottom / side so they just touched, with a cone sitting in the dead spot in the centre it would probably be as optimal as it could be made with hand tools (short of custom molded plasted)

yep, a 3D printer could also be used to make central nozzle.

Yes those are really cool, I saw a motor bike chain done with one of those and all the links moved like a real chain. That would be perfect to create a five sided scolloped cone.

Originally Posted by HeadScratcher

Yes those are really cool, I saw a motor bike chain done with one of those and all the links moved like a real chain. That would be perfect to create a five sided scolloped cone.

We have one here at work - the surfaces it makes are not that smooth but a bit of polishing would get it right. I'm not allowed to touch it at the moment as the guys that bought it are pretty possessive about it.

Tell them to share their toys

Did a mock up of 5 x 90 mm (45mm rad OD) pipes at 45 degrees into a 150mm pipe (75mm rad OD).

My maths says it should fit with 5mm between each pipe but for some strange reason there seems to be an interference fit.

Managed to clean up the picture after much mucking around. Still at a loss to understand why the math didn't pan out...

Worked out the flaw in my math… the circumference of the 150mm pipe may be 471mm, but what I am effectively doing is joining 5 straight lines in the form of a hexagon. As the diameter of the joining pipes are measured in a straight line and measure 90mm I really should have calculated the hexagonal circumference of the 150mm pipe.

Can I get away with only 4 x 150mm pipes? Otherwise I would need to make the collection area 200mm (and need to buy a separate length of 200mm pipe) and then step down to 150mm. In the interest of keeping the project simple 4 x 150mm pipes would be more practical.

Nominal ID of "200mm" pipe is 225mm.

Thanks Chris, as you can see from the attached pic the parimeter of the red pentagon is much smaller than the circumference of the circle, and since the joining pipes are measured across the flat, I should have been working off the outside measurement of the pentagon. Which explains why my pipes overlap and there is no gap.

Originally Posted by Chris Parks

Nominal ID of "200mm" pipe is 225mm.

Hi Chris can you verify this measurement please, do you mean 195mm?

Welcome to Iplex Pipelines Australia Pty Limited