Water Blaster Water Flow Overview

[isoaker]

This article on water blaster water flow took a little longer to put together than I first expected. And even so, I know I could easily pour many, MANY more hours into if were I to try to draw and explain things with more accuracy.

That said, while accuracy may not be there, the general concepts should be ok and some may find it interesting to spend a little more time thinking about exactly how water ends up flowing through these systems we use.

Optimizing water flow is most important for stream generation, though flow during the pumping and pressurization process is important as well. Having a better understanding and appreciation of water flow is particularly important to those who want to make improvements to their water blaster's performance.

[SSCBen]

Didn't notice this until now. Overall, this is pretty good.

I suggest noting a few other sources of turbulence: pipe roughness, slight imperfections in any geometry, and vibration from moving a water gun around. Experiments have shown that if you create as perfect a pipe as possible and isolate it from vibration, you can delay the transition to turbulent flow to essentially an arbitrary degree.

I might suggest eliminating the word "lamination". It is not used in the scientific literature, and is vague. Can you measure lamination? The answer is no. I might suggest "turbulence intensity" as an alternative, as this can be measured. Low turbulence intensity is what we want. Though this term might not be for the audience you are shooting for.

Another thing worth noting is that the velocity at the walls is not just slower; it's zero. We're not dealing with micro-scale pipes, so there's negligible slip at the pipe walls. So "While the distance from the middle to wall of the tube does not experience as drastic a speed reduction as the above illustration shows, the effect is real." is not accurate. The transition section is called the boundary layer and it is generally very thin. Thus, approximating the average flow as a flat profile in a straight pipe section isolated from other elements (e.g., upstream bends, etc.) is usually okay for some problems.

On the other hand, length is used to denote the speed of flow. Thus, longer arrows mean that water in that region of the tubing is flowing more rapidly than areas with shorter arrows. The smallest arrows that are draw curling beyond 180 degrees back are meant to illustrate drag on the higher stream flow and not meant to suggest that water is actually flowing backwards.

Turbulent flows are usually analyzed by breaking the flow down into a time average flow and turbulent fluctuation. The turbulent fluctuations could easily cause the overall local flow to go backwards for a short period of time. The small arrows are reasonably accurate representations of the turbulent fluctuations, and I think you should label them as such.

One solution to improve overall lamination through wider tubing is to insert a flow lamination device (a.k.a. Laminator) which serves to re-equilibrate flow rates across the entire cross-section of the wide tubing. Laminators work on a rather simple principle: insert a matrix of smaller diameter tubing into the flow path for a short distance. Since each piece of smaller tubing can only accept a percentage of the overall water flow through the larger tubing, the central flow from the wide tubing ends up needing to spread out across several of the lamination tubes in order to pass. This slows down the central flow while the outward pressure from the central flow serves to push against and increase the speed of flow in the outer tubes in the laminator. Thus, while the central flow rate is partially slowed, the outer flow rates are partially increased, bringing the resulting flow after the laminator much more synchronized, thus much more laminated. Turbulence is also constrained simply by the fact of reducing the area (volume) available for the flowing water to create turbulence. There is a very interesting (and complicated) interplay of forces going on with water passing through a laminator. Perhaps one of the most famous water blaster, the Super Soaker CPS 2000, features a lamination device just after its pull valve, right before its nozzle. This is one of the reasons the stream from the CPS 2000 flows so smoothly out the nozzle, giving this water blaster such impressive ranges.

This is not the right explanation. I don't think the laminator* does much to flatten out the velocity distribution, and if it did, that wouldn't necessarily be helpful. A laminator works by preventing flow in the direction perpendicular to the pipe axis and increasing dissipation of turbulent energy (from the lower local Reynolds number). This reduces turbulent velocity fluctuations (especially in the direction perpendicular to the pipe axis), and it also reduces "swirl", which is a mean (i.e., non-turbulent) motion around the pipe's cross section (i.e., like a clock hand). Swirl is very bad for stream cohesion, as is heavy turbulence. Some sprinkler nozzles induce swirl by adding a corkscrew, causing the water stream to break into droplets immediately upon exiting the nozzle. I can explain more, or in different ways, if you would like.

* I haven't seen this term used in the scientific literature either, though "turbulence control member" sounds pretty lame, so this is okay. "Honeycomb" is also a common term.

[isoaker]

Thanks for pointing out some problems and some definite mistakes in understanding!

I'll work on making some adjustments to the article based on this very helpful feedback! However, it'll probably take me a few days to adjust things completely.

It'll take time to replace "lamination" throughout the article. And while conceptually considering flow speed at the wall as being zero, somehow I cannot accept that as true in practice. Were flow speed at the wall zero, a wet tube could never be emptied or flushed completely. At some point, if a tube were continually being flushed with water, the water molecules originally within the tube should be completely replaced by new ones. I can accept the idea of flow by the walls to be slow enough to be considered approximately zero, but an absolute zero flow doesn't feel right.

[SSCBen]

And while conceptually considering flow speed at the wall as being zero, somehow I cannot accept that as true in practice. Were flow speed at the wall zero, a wet tube could never be emptied or flushed completely. At some point, if a tube were continually being flushed with water, the water molecules originally within the tube should be completely replaced by new ones. I can accept the idea of flow by the walls to be slow enough to be considered approximately zero, but an absolute zero flow doesn't feel right.

You are right. The no-slip condition is often confusing at first and it is only an approximation. It is an excellent approximation in most practical situations (you can read Wikipedia for a short summary of when and why it does not apply). Here's a video showing an experiment with colored dye. You can see that the dye nearest to the wall does not noticeably move.

Your thought experiment about emptying a pipe is interesting, and can help illustrate some of the approximations involved in fluid dynamics. Here's one way to resolve it: in the model (but not reality), the fluid directly at the wall has an infinitely small (i.e., zero) volume, so it never needs to be removed. The no-slip condition mainly applies to flows where the continuum approximation holds. This approximation basically ignores the discrete molecular nature of the fluid because the model averages the motion over local molecules. In reality, individual fluid molecules will be moving near the wall. But if you average all of them near the wall (as the continuum approximation does), the average will be very nearly zero.

Simulation techniques that look at individual molecules do not impose a no-slip condition, rather, they actually model interactions with the wall. Those techniques are becoming more popular as computers become more powerful. I am taking a class on them next semester.

Let me know if any of this doesn't make any sense. I'd really like to see introductory articles on water gun fluid flow that are as technically accurate as possible.

Edit: Another point worth mentioning is that the no-slip condition only applies in the direction of the wall. So the fluid can be removed from the wall by moving perpendicular to the wall.

[isoaker]

Thanks for the further clarification!

Also, thanks for sharing some of your knowledge on fluid dynamics with the group! I definitely appreciate it and am learning (and learning about what I need to learn more about ) from the great info you've posted here and in the other thread. Rigid body physics is so much simpler than fluid dynamics, but fluid dynamics is far cooler, IMO. *pun intended as well*