Page 1 of 1

Laminar Flow

Posted: Fri Aug 19, 2011 7:52 pm
by atvan
Laminar flow is the best choice in all situations other than a riot blast, right? Wrong. Turbulent flow has advantages as well. It our situation, it usually tends to hinder more then help however.

Laminat flow is the flow of a fluid devided into destinct laminae, or layers. These layers slide past each other with minimal friction. The movement between these laminae is nonexistent in true laminar flow. Now, for us, this is ideal. However, it is very difficult to acheive. We can, however, utalize the properties of laminar flow if we do obtain it. In a ball valve, for example, flow is often restricted due to the design. If the valve slopes into the hole, the small hole will experience a loss of pressure, and an increase of velocity. So even though the hole is small, it need not restrict flow at all.

Ben is always telling us to avoid bends in the piping. This, however, is especially important farther from the PC, where the pipe is usually straiter anyway. Don't add unnescesary bends, but having some may not cause turbulent flow.

The reason for this is the Reynold's number or Re. The Re is defined as the ratio of density to viscosity or the fluid, multiplied by velocity and distance. For our water guns, the density is roughly 1, as is the viscosity at room temperature. Thus, in all practical terms, the Re for our water is velocity*length, unless I got my units messed up. The bends in pipe increase velocity, which is why they contribute to the turn to turbulent flow. The Re for 10 cm of pipe carrying water at 100 cm/s then is 1000. For an average CPH, then, the Re is about 2000. This number is low enough to maintain laminar flow in a fairly smooth pipe, such as what we use. Later on in the path, the fittings may be enough to intiate the transition to turbulent flow. However, if a nozzle accelerates the water to 10 m/s, or 1000 cm/s, the transition to turbulent flow is complete at a distance of just 4 cm, due to the fact that the critical Re an less due to drag and wind, etc. This does not mean that your stream starts breakup there, just that the laminae have disapated and the stream is now triying to tear itself apart, with inertia now prevailing over viscosity.

Turbulent flows have higher drag, but also higher energy. Laminar flows contain a thin slow boundry layer, where a turbulent flow has a thick fast boundary layer. This enables turbulent flows to better avoid separation bubbles, which don't really matter for us. However, they have a higher energy, which is useful. Also, the faster moving bounary layer means less of that annoying drip from the stream in some cases, though the edges are often sheared off.

In essence, laminar flows have range, while turbulent flows have better stream speed. A turbulent flow will also induce higher output from a nozzle. This knowledge may prove useful in designing nozzles and guns for the future

Also, for those of you who read my entire rant, thank you. For those of you that did not, that's OK too.

Re: Laminar Flow

Posted: Sun Aug 21, 2011 2:43 pm
by atvan
Anybody? I thought this might spark some interest from at least somebody. Maybe it is the community war backfiring. If you have questions, I'd love to answer them.

Re: Laminar Flow

Posted: Sun Aug 21, 2011 3:10 pm
by Shrike
regular tubes achieve res of about 2300. but with special arrangements, you can get much higher res while still maintaining laminar flow.

Re: Laminar Flow

Posted: Sun Aug 21, 2011 6:10 pm
by atvan
Yes, but even a slight imperfection will cqtylize the transition.

Re: Laminar Flow

Posted: Mon Aug 22, 2011 2:32 am
by C-A_99
Increased velocity can be helpful at the expense of some lamination, but there seems to be little testing on this done so far, and is likely very tricky to gauge.

Re: Laminar Flow

Posted: Mon Aug 22, 2011 2:53 am
by SSCBen
The "critical Reynolds number" business really only applies for straight pipes. Re = 2300 is about where imperfections in the pipe will cause turbulent motions. If the pipe was perfectly smooth, there would be no turbulence in this case.

Bends change this completely. The flow coming out of a bend is definitely turbulent to some degree. Is this bad? Maybe not. Test it and see.

If you understand the definition of the Reynolds number, then you'll understand why glycerin helps range (it increases the viscosity, which decreases the Reynolds number) and why laminators work (they decrease the characteristic length, which is the diameter here---straws basically keep the fluid from moving in directions other than down their length).

If I had more time I would happily detail a lot of what I've learned since I wrote a lot on SSC, but I unfortunately do not have the time.

A lot of research has been done into the atomization process (i.e. the breaking up of the stream) and we almost certainly could use some of this research to figure out other ways to improve the range of water guns. Try using Google Scholar to look for some technical papers if you are interested in this.

Re: Laminar Flow

Posted: Mon Aug 22, 2011 2:55 am
by atvan
Ben, the bends are an increase in velocity, hence their effect. Also, viscosity raises the Re, but also the critical Re.

Re: Laminar Flow

Posted: Tue Aug 23, 2011 2:52 am
by SSCBen
Bends don't change average velocity appreciably. Turbulence is not determined solely by the average Reynolds number. But the Reynolds number can explain a lot. The problem with bends is their geometry, which creates eddies. Add some imperfections to the walls and you get turbulent flows.

Also, viscosity does not change the critical Reynolds number for pipe flows. The Reynolds number itself certainly can change, but the critical Reynolds number in this case seems to be a constant. That is, after all, one of the reasons the Reynolds number is used. (I think the Mach number of the pipe flow might affect it, but that's irrelevant here and probably not well studied.)