Minimum velocity to keep particles in suspension

[engineeratfit]

Hello,

I am trying to design a dilution system that would have a 1-10% particulate in fluid. This is spinning in a loop (vertical and horizontal components) and I am trying to have it fairly homogenized so that I can pull a representative sample.

I have used Stokes Law to find a terminal velocity for the fluid/particulate pair. My understanding is this is the point where the particles no longer accelerates but still moves at the calculated velocity. Do I simply need a velocity greater than the terminal velocity? How do I determine the velocity to circulate?

 

[bimr]

You don't provide much information to go on. The minimum velocity depends primary on the type and size of the solid. Typical minimum velocities are 3 ft/sec for horizontal pipelines and 4 ft/sec for vertical pipelines.

http://www.engineeringtoolbox.com/slurry-transport-velocity-d_236.html

 

[MikeHalloran]

You need to read several chapters in Govier/Aziz " The Flow of Complex Mixtures in Pipes".

Allocate substantial time.



Mike Halloran
Pembroke Pines, FL, USA

 

[Artisi]

What is the material being pumped, sizes etc?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[davefitz]

I would suggest the fluidization text "fluidization and fluid particle systems" ( 1960) by Frederick Zenz. One would need to exceed the terminal v3elocity determined by Stokes formula to account for velocity profiles in the pipe and acceleration that occurs at elbows, tees, wyes, etc.

"Whom the gods would destroy, they first make mad "

 

[engineeratfit]

Thanks for your responses. Particle size is uncertain at this time. I am trying to make an excel spreadsheet to determine the velocity needed for any given size. The expected range is 100um to 1mm. The particulate is very dense (4000+ kg/m^3) and it is in diesel fuel. Again, this isn't really a slurry but instead a small amount of it in suspension.

It seems that some people say to calculate Stokes and exceed the terminal velocity. Others seem to point to something much faster and more complex. I will look into those two books and I appreciate the recommendations. I've been able to avoid fluid dynamics since college and it is now coming to haunt me!

 

[racookpe1978]

Think always about how "accurate" you need to be:
If you miss 10% of your biggest possible particle, it that OK because you can collect 90% of the particles?
Or do you need to suspend 99.99% of those maximum sized particles so you know you are not letting any stay out of solution?
If you only suspend 10% of your maximum sized particle, is that OK because your sampling is still sufficient to tell you are "some" of the particles are present?

 

[btrueblood]

I think the Stokes number gets you a minimum velocity to cause transport (movement) of the particle (e.g. in a stream bed); but realize that you will likely see significant stratification of the particles by size/density at lower velocities (this is how sluice boxes work). You mentioned sampling, which sounds like you want to get a good idea of the relative size distribution of the particles - then yes, this implies using much higher velocities, or even mechanical stirring methods, in order to ensure that all particles are reasonably randomly distributed throughout the fluid.

 

[engineeratfit]

You are correct. This is intended for pulling samples for size/distribution measurements. I want to see a correct distribution of particle sizes.

 

[geesamand]

I think you need to consult with mixing literature. They refer to this as a Solid Suspension problem.

(Whether or not you need to include a mixer is not the point)

 

[davefitz]

as per <

http://www.particles.org.uk/particle_technology_book/chapter_9.pdf>

the pneumatic conveying guidelines are to use 150% of "saltation velocity" to ensure uniform flow.

"Whom the gods would destroy, they first make mad "

 

[racookpe1978]

If you have any velocity at all, and only need to stir the mixture to get an "adequate" sample - and perfectly mix the liquid up, I'd insert a fixed mixer (like the criss-crossed plates that automatically stir epoxy resins together in the dispenser ahead of the sample port. Then have the sample port be inserted slightly up into the pipe towards the centerline like a pitot tube so it doesn't get clogged with solids on the floor of the pipe.

the fixed plate mixer stirs the liquid and solids to agitate them slightly at low flow loss, and the sampling tube extracts the mix at a known height so you don't always get junk from the bottom of the pipe.