Hacker News new | past | comments | ask | show | jobs | submit login
Why does a spray bottle work? (alexanderellis.github.io)
601 points by otras on June 9, 2018 | hide | past | favorite | 123 comments

A few years ago, a British inventor appeared on a TV show called Dragon's Den (UK version of Shark Tank) and pitched his idea of an improved spray bottle called the Anyway Spray.

Instead of a normal plastic tube inside a bottle with an opening at the end, he designed a tube with millions of tiny perforations along its length. This meant the spray bottle could operate at any angle (even upside down) and could empty all liquid in a bottle with no leftover.

Here is a 1 min demo of the spray: https://vimeo.com/195318537

Here's an explanation of how it works https://www.youtube.com/watch?v=xXk46rvzdpY

And here's the original pitch he made on Dragon's Den https://www.youtube.com/watch?v=RjOvyjeZyIw

If it was simply perforations, it wouldn't work when the bottle is upright because those perforations would let the air into the top of the straw. Rather, the straw is made of a membrane which lets water through while blocking gas.

I’m not sure you can make a membrane that does that? I get the sense there is a membrane that wicks the fluid around the surface of the tube so you always have to pull in liquid before the gas? Maybe that’s the same thing.

You can retrofit it into an existing spray bottle with some flexible tubing and something to weight down the end, like some stainless-steel nuts. That way, the end of the tube follows gravity along with the liquid and ensure it remains submerged.

I don't think there is any idea here other than retrofitting this innovative straw into existing spray bottles.

You need to let air through or else it creates a vacuum. What's happening is it just slows the flow of air enough to be able to get suction from the fluid. Having fluid in the tube also helps.

You need to let air through into the bottle to replace the volume of ejected water. No air should ideally be entering the straw.

Seems like he managed to license it to Tesco's range of label products. Good for him.

Good for us! :)

Why isn't this in everything? I'd easily pay an extra few cents to have this in spray bottles. It's such a frustrating problem.

In my quick analysis (could be mistaken, I welcome corrections), if this is one-time disposible then it only makes sense if the value to the consumer of the amount of product not thrown out as a result, is less than the extra cost of licensing and manufacturing this and this cost can be passed onto the consumer somehow.

In other words, if this costs 5¢ more then it is worth spending an extra 5¢ on if you keep from throwing at least 5¢ of product out. Below this it is better to throw a little product away. (However, consumers might overvalue being able to use the last little bit of product. Perhaps they value at 25¢ to not throw the last 1/30th of the product away, even though they paid less than $7.50 for the whole package. i.e. perhaps they paid $5.00 but would pay $5.25 to keep from throwing away the last 1/30th of the product. However this seems unlikely!)

More likely, if it costs 25¢ it is more economical to throw away, for example, 5¢ of product instead, by not having it.

However this analysis assumes that the 25¢ is passed onto the consumer. If the consumer is very price-sensitive (for example there is a standard price and very standard volume, as is the case with milk, though that isn't dispensed this way) then adding this as a competitive advantage would benefit the consumer, while the supplier eats the costs as a cost of acquiring this consumer.

This is my quick analysis.

It's nice that it would end up using all the product but I'm thinking more about convenience. I often hold a bottle sideways and, while there's plenty of liquid, I can't get the spray to work in that orientation. So frustrating.

Here's a cheaper and faster hack: soft tube + weighted end.

The weighted end will tend to go towards where the liquids are. Imagine a bottle being held lengthwise. The weighted end will be in the same location as where the liquid is.

Turn the bottle upside down... the weighted end drops to where the top of the bottle is, along with where the water is.

Standard pressure thingies still work - you still get a spray. No need for custom membranse.

What it really comes down to is the question of the actual goal and the costs of reaching that goal. For a huge percentage of things "good enough" is in fact good enough. It's also important to not confuse the goal and proposed solutions for reaching that goal.

For the example above, is the goal to be able to spray upside down? To ensure that the sprayer is able to spray a higher percentage of what's in the bottle? How large is that percentage improvement (90% up to 95%?), what's the incremental cost of increasing that percentage, and what is the actual worth of 5% of the contents of the bottle?

A draw tube that is flexible enough to fold in half, strong enough to not kink, sturdy enough to not collapse when bursts of suction are applied, weighted enough to stay in liquid, strong enough to assemble quickly and easily, etc. is probably also expensive enough to just not be an economical solution.

By the way, if you don't purchase cleaning supplies regularly you may have missed that a lot of those tubes have gone away - they're now molded as part of the bottle so just tipping the bottle will put all remaining liquid at the intake point.

I saw an instructional thing about that. They did it with a bit of silicone tubing the same diameter as the original, and a steel nut with matching inside diameter. Swap out the tube and screw the nut over the end.

I'm not sure you could actually make it work upside-down without crimping, but at any normal usage angle it works great.

What cheap alternatives to small bits of metal could be used for the weight?

Metal could interact with and damage a lot of different things.

the metal doesn’t need to be exposed to the contents in order to serve its purpose. It just needs to be on location.. could it not be enclosed by the same material of the straw?

Downside: it's harder (read more expensive) to prevent a soft tube from collapsing under the vacuum.

Jumping to the conclusion that everything a company does is calculated and delibarate is a mistake I see a lot of people make, but in this case it’s accurate.

In a business that operates at a scale of 100s of thousands if not millions of bottles sold annually, cents matter and can make up a significant portion of revenue.

With this in mind, there is a huge incentive to get people to buy sooner rather than later and more often. Forcing a small amount to remain in a used up bottle is definitely a benefit to the company even if the product is going to waste. This doesn’t include the cost of implementing something new. I’d imagine a company would do extensive survey based research to understand if the additional customer satisfaction would outweigh the costs/downsides.

If it’s not a conscious decision, it’s a nice coincidental benefit to them.

I am not sure why you were downvoted. There are several examples of the behavior you described.

Potato chip companies reducing the quantity of chips in a bag but keeping the size and price same. Because they were unable to raise prices with older packaging. Typically customers didn’t look at weight, just size of package. It was in the news for quite a while and IIRC there is a business case study.

It has been happening in beverages and other items too. Why do you see now package weights of odd numbers like 11.5 or 14.5 oz instead of standard number like 12 oz or 16 oz with typically no price reduction.

Recently I encountered same behavior with toothpaste tube in Japan. To keep the form/shape of the toothpaste tube with usage the nozzle was designed to suck in air to fill new space in the tube. My wife claimed we are out of toothpaste and wanted to throw away the tube. The toothpaste tube was perfectly formed but I noticed tube was still little heavier than I would expect for a empty plastic tube. But when you squeeze the tube mostly Air was coming out with little bit of spray of paste. Reducing the air space with paper clip, we were able to use paste from same tube for another 2 weeks. Typical behavior would be most probably what my wife was going to do to throwaway old one and open a new one.

Maybe, but it seems more likely this isn't due to malice but just stupidity/apathy. Maybe most companies just aren't aware of this technology, or they don't want to implement something new because of uncertainty about the potential problems with the new technology. For example, maybe 1 out of a 100 of these bottles gets completely clogged and won't dispense anything at all. That looks bad for the company and maybe even requires a recall. It's easier to just stick with what they are doing and no-one is complaining loudly about the small bit of wasted product.

Large companies are constantly doing research and development which handles all of this. They definitely aren't stupid and aren't waiting around for youtube videos to learn about things.

They have teams that actively engage with new suppliers and inventors and are also getting pitched constantly with new ideas. It's most likely that extensive testing and consideration showed that this wasn't worth the costs. You even described such a scenario so I'm not sure why you think that's stupid or apathetic, when it's actually quite calculated.


Business benefit supersedes customer benefit. Always.

I wonder how the economic and environmental costs of the extra tubing (the tube in Anyway Spray bottles is longer than a straight, conventional one) compares to the costs of those few last drops for most sprayed products...

This improvement would cause the manufacturer to sell less product by volume.

I noticed the video described him as MBE, so I looked him up. He also invented a water filter and was named Member of the British Empire for it: https://ideas.ted.com/since-the-ted-talk-michael-pritchard-o...

That episode of Dragons Den is infuriating. Here is a guy who has invented something with a widespread, common appeal. He invented a product that potentially fixes a real problem in 1,000,000,000 units per year and he has to give away 20% of his invention to get support.

The investors on Dragon's Den always want too much equity - I would imagine that anyone going on the show would know that.

I also think some people go on the show knowing that they're not going to get investment at a sensible equity level, and are really on there as a way of marketing their idea to a wider audience.

He had the option to reject the deal, no? It's a free market for venture capital, so if that deal is so bad and the opportunity is so good, he should have been able to find a better one.

From my perspective, having a great idea is not enough. Even having a proof of concept is not enough. There's a lot more that goes into building a successful business. Investors have to price in the chances of all the things that can go wrong in the process against the perceived value of the product.

Likewise, as a founder, you often need at least some investors that can help propel you forward in some way. That can be worth a lot more than dumb money.

I haven't watched the episode you mention, but generally this is how I think about it.

>Likewise, as a founder, you often need at least some investors that can help propel you forward in some way.

Which is the infuriating part, when the "investors" contribute nothing to the project but a tiny slice of money they didn't work for to begin with, and get vastly disproportional returns for it.

I feel like you've missed my point.

There are a limited number of good opportunities to invest in, so it's not like investors get to unilaterally set the terms of these deals. They have to compete with each other, at least on the most attractive deals. A savvy founder turns down money on bad terms.

I got your point, I just don't agree with it! I think that's a consistent logical position to take, but I don't think it's undeniably _ethical_ because of that.

I guess I'm missing what it is to disagree with. Are you arguing that people with good ideas should simply be entitled to the capital to pull them off, to avoid the need to compete in capital markets?

I mean, sure, that'd be great. I think it'd need one heck of a societal overhaul to achieve, though.

If we wanna keep it in terms of a capitalist market, the problem is that investors are essentially selling a service just like anyone else, but because wealth inequality is where it is, there's relatively little competition in that "market". Thus, what you get is very high "prices" for relatively poor "service". This is a problem.

In an ideal free market, investors are just another kind of business, and would compete both on price and their offering such that everyone was getting a good deal and everything moved along efficiently. In the real world, investment is a very concentrated market, meaning that the deals are shoddy and inefficient.

Welcome to venture capital.

That it is common makes it neither more nor less just.

Yes, exactly.

I see your point and I appreciate your perspective however, he didn't just have a proof of concept, he had a product. It is completed and working.

Based off of a Daily Mail article about it [1], it looks like the deal fell through anyways, so (hopefully) he was able to retain full ownership in the deal with Tesco. I love the idea and wish Mr. Pritchard the best. Hopefully they can get it over to the US soon!

[1] http://www.dailymail.co.uk/sciencetech/article-4335468/Briti...

I have a spray bottle that I bought to clean windows, because it can shoot a stream about 20 meters which is great. It is also labeled "sprays upside-down!" which is really misleading. I thought it sprayed in any orientation but it doesn't. It only sprays when held straight up or straight down. There's an inlet on the base of the handle for the inverted case, and a valve that prevents the main pickup from drawing in air.

That’s not “really misleading,” it’s “barely misleading” or perhaps “slightly misleading.”

I think most people would assume the marketing meant "can spray at any angle, even the hardest case of upside down".

I knew I did when reading this thread until I saw this comment.

To be clear, there are two products: one labeled “anyway spray” which sprays in any direction and one labeled “sprays upside down” which sprays right side up or completely upside down. Seems fair to me.

Considering the replies this has gathered, I'll add this (half-obvious) anecdata FWIW: you can, in I'd say 70-85% of cases, recover all of the product in a spray or pump if you don't mind (potentially repeatedly) fiddling with the hose.

Just pull the hose out and bend/twist it really hard in a specific direction. If it's not utterly stiff, winding it tightly around a pencil (or your finger, or just twisting the hose to that tight of a bend radius) is ideal. Hopefully this much flexing means that after naturally uncurling the curl will still "stick" to a useful degree for longer than a single usage.

One non-obvious extra detail: if the product is in a non-transparent spray bottle, before replacing the lid, rotate the spray pump so it faces the direction you've bent/curled the hose. Now a) you know which direction the hose is pointing and can eliminate "where is it in the bottle", and b) you just have to face it forward for it to work, which is intuitive.

I know it's for TV but that was a great pitch.

Cool. Looks like the guy got his stuff into products. He got raped by the investors in my view.


This is a nice example to show the illusion of explanatory depth [0]. Most people would say they know how a spray bottle works, but fail to describe all relevant details of the implementation,

[0] https://www.edge.org/response-detail/27117

> If you asked one hundred people on the street if they understand how a refrigerator works, most would respond, yes, they do. But ask them to then produce a detailed, step-by-step explanation of how exactly a refrigerator works and you would likely hear silence or stammering.

This is a terrible example because its essentially a trick question. The illustration would highly depend on context.

Say you meet someone on the street and you casually ask them this question, or you ask this on a birthday. "Works" with regards to a refrigerator refers to "knowing how to operate", or perhaps the basics.

Whereas if its for the TV quiz program "engineering of the kitchen" or "an applicant for technician repairing Samsung's refrigerators", yes then its clear that more depth is meant.

That's not to say the Dunning-Kruger effect doesn't apply here. Its just that there's a level set for a debate due to the context which then presets also the level of the expected answer.

Michael Faraday gave a great series of lectures on how candles work that also illustrates the difference between cursory knowledge and explanatory depth. Most people think they know how candles work, but when you dive into the details there are all sorts of interesting things going on.


(you can also buy the lectures in book form on Amazon.)

Engineer Guy (a fantastic YouTube channel for learning how things work) has an audiovisual curriculum for this book.


That is indeed a wonderful recreation of a brilliant original, and I often think of it when trying to figure out how to provably and clearly demonstrate some absolutely novel concept. Faraday's genius shows through here.

Thank you for sharing this!

Having watched the first lecture, I was thoroughly impressed by the design and presentation of this historical artifact, with (judiciously chosen) modern embellishments.

There is a relevant interview [0] with Feynman that I always remember.

[0] https://www.lesswrong.com/posts/W9rJv26sxs4g2B9bL/transcript...

I think how trains stay on tracks [0] is also a good example of this.

[0] https://www.popularmechanics.com/science/a25581/science-behi...

Yes! I coincidentally just learned about that topic recently.

What’s great about that problem is that it isn’t obvious that there even is a problem in the first place. “How do trains stay on tracks? Well... they have wheels, don’t they?”

Then when the problem is explained, it sounds impossible. The solution you first think of is probably wrong.

One way valves are basically diodes, but for fluids rather than electrons. You can also make logic gates out of just diodes (or one way valves).

Imagine writing software for a hydraulic computer!

MONIAC [1] was an analog computer designed in 1949 that used water to model the national economy of the United Kingdom.

1. https://en.wikipedia.org/wiki/MONIAC

Made by an engineer turned economist from New Zealand no less.

Sadly he is better know for a diagram he drew of an observation, the Phillips curve.

I fear that much like this hydraulic model, Steve Keen's work is likely to be ignored or ridiculed by future economists while they fret over religious X diagrams.

Hydraulic models solve differential equations. They aren't bad , but modern computers are a much better computational environment.

There's a fairly significant overlap between engineers and economists. Chemists and physicists also.

Largely out of the economic mainstream, since the 20th century, sadly.

This was brilliantly parodied in Terry Pratchett's Making Money.

I got a Kindle for Christmas, and I've been reading through Discworld chronologically. I'm on Raising Steam right now, and I already feel like I'm going to have to reread the series with a list of references open. There were a lot of places where it was apparent there was a reference to something, but it was completely inaccessible as an American in his late 20's.

At the same time, I almost feel like I shouldn't read the last book for a while, it's going to be really bittersweet.

Do you know about the Annotated Pratchett File? https://www.lspace.org/books/apf/

I luckily don't have to imagine it. During an electric course in school for half a year where we learned all kinds of stuff like high voltage circuitry, we also learned pneumatic logic.

Almost all electric components can be replicated in pneumatics; capacitors, resistors, transistors, diodes (including Zener diodes)... And from that you can form simple logics.

One design we had to generate was to build a pneumatic gate which would open and close by pressing four buttons (two for each side of the gate) and light up a red, yellow or green light depending on whether or not the gate is moving (of course the lights use a pressure sensor).

Of course nobody builds that because such a system would be expensive compared to a Attiny with a shift register and a few relays capable of doing the same.

If you've ever wondered what powered the logic of a hydraulic transmission back in the olden days, it's essentially a hydraulic computer with all sorts of components which are analogous to diodes, transistors, even a pump-driven comparator to weigh input shaft speed and load against the output shaft to decide when to switch a gear.

Do an image search for an automatic transmission valve body. It even looks like a big industrial circuit board. Some components are passive like a check valve, some are active like a valve which is operated by pressure from some other part of the circuit. Newer ones also have partial electronic control, with microcontroller-driven solenoids and actuators moving around valves to control hydraulic rams which finally engage or disengage clutches. But it all began with no electronics at all. Pretty neat trick if you ask me.

I remember being a kid taking apart my first Ford transaxle and being amazed by the sheer complexity, almost to the point where it seems biological. My dad bought a really bad one for a couple bucks at a junkyard. We didn't need it to fix anything, he just wanted to feed my curiosity. It was a great parenting move. I recommend doing it yourself if you find a cheap busted one, just to get the ideas flowing.

> One way valves are basically diodes, but for fluids rather than electrons.

The analog of the fluid in that analogy is electric charge/current, not electrons which move in the other direction (which kinda breaks the analogy as a result).

However, I think it would have been quite difficult for me to get the concept of "holes" in semiconductor current, which are positively charged pseudo-particles created by the absence of electrons, if I hadn't been able to think of them as being like bubbles in a fluid. They go the opposite way of the fluid flow, and they are somewhat particle-like. So it's not a totally broken analogy, it does help to clarify some things.

I generally found the fluid analogy did more damage than good for a lot of people in EE school. A lot of people think of electricity as electrons zipping around a circuit at the speed of light due (at least in part) to the water analogy. Having to transition from that world to thinking in charge transfer and fields while electrons move around linearly at a snails pace if at all was very difficult for many people.

I'd bet in a poll of the general public, most people think electrons fly around a circuit like water in a pipe, which is of course completely wrong. I guess it's debatable whether the inaccuracy matters or not. But in general, I prefer analogies that are simplifications and leave out details, rather than those that are completely ass backwards from reality.

That's an interesting experience. I'd say that the idea people have about water in pipes is probably overly simplistic as well, and an analogy to an equally poorly understand concept can't help very much. To pick up your example about electron speed vs. signal speed: Water pressure is transmitted at the speed of sound, which is much, much higher than the speed of water flow -- water taps wouldn't work otherwise.

The way the analogy was explained to me when i was a student is that the conductor is a hose filled with water, when you open the valve the water comes out instantly at the other end because the "charge" is moving. It was never confusing to me.

This is the correct way. In the water analogy, the circuit is a closed system and components are completely filled with water at all times. For example, the capacitor is a filled sphere with a rubber divider. The total amount of charge/water is constant, but you can change how it is divided between the two plates by an applied voltage/pressure.

The analogy is valid and hydraulic circuits work.

If people transfer their incorrect understanding of water to incorrect knowledge of electricity, that doesn't make the analogy bad.

I think anyone that has waited for hot water to get to the tap understands this. Pressure (voltage) transmits at the speed of sound (light). The speed of the particles through the medium, on the other hand, depends on many factors.

I think the analogy can work, but needs to be carefully applied. This wasn't an analogy taught in my EE program, but one that is common in TV programming and provided by high school science teachers whom may not themselves understand the nuance so many came into the program with this understanding in place.

Ultimately though, this concept was the breaking point for my EE class; It's when half the class dropped out to other majors. Not that I blame the analogy entirely for this, I think in general this is the point at which a lot of people lose connection with deeper sciences. It's the point where analogies to the physical world (matter) break down because we're now fundamentally talking about energy, which doesn't obey the rules all humans are accustom to. The world where we can't see it or relate it to our daily experiences but only talk about it through mathematics.

And also move really really slowly compared to the field. Their (net) movement is actually pretty irrelevant to the propagation of what we call electricity an sich.

There was a post here on Fluidics a few weeks ago.

A very simple design, based on a principle that was discovered over 2000 years ago.

http://www.mdpi.com/2073-4441/7/9/5031/pdf (see page 13-14)


That diagram at the top of the Wikipedia article made the concept more clear to me than the whole orginal post. Drawing the valves like that really helps.

Edit: grammar

That's funny, I don't know what that diagram is trying to convey.

Your confusion is understandable, since the Wikipedia article doesn't even make the slightest attempt to explain the diagram.

So let me try. Let's start with the piston pump on the left. The brown piece moves up or down in a tube, and the two black "T" figures are valves that open or close in response to water pressure.

The down arrow next to the piston indicates that the piston is currently moving down and pushing on the water. This causes the valve at the bottom to close and the valve on the left to open. So water comes out at the top left.

After the piston is pushed down as far as it will go, you start pulling the piston back up. This is not shown in the diagram, but now everything is reversed. The valve at the left closes, and the valve at the bottom opens, both in response to the pressure change.

So now, water is pulled up from the well or other source, filling the chamber. When it is full, you start pushing down again, and the cycle begins anew: the valve on the left opens and the valve on the bottom closes, pushing water out of the spout at the top left.

Once that makes sense, take a look at the plunger pump on the right side of the diagram. You will see that it works on exactly the same principle: when you move the brown part up or down, it pulls or pushes on the water in the same way as the piston pump. The only difference is in the shape of the moving part.

The most recent AvE video had a break down of a pressure washer which includes an explanation of a piston pump. https://youtu.be/4IbjQyR9X6E

Oh, ha. Thanks for the explanation. Yesterday, I understood the piston diagram, and I was scratching my head looking for the difference between the two. Good to know I wasn't actually missing something.

According to the plunger pump page, there is another practical difference - the pressure seal on the plunger pump is "stationary" allowing for higher pressure. Not sure why being stationary relative to the housing (but not relative to the plunger) makes the difference there.

> Your confusion is understandable, since the Wikipedia article doesn't even make the slightest attempt to explain the diagram.

A little meta, but in these situations I usually try to improve the source and post a reply saying I edited it, instead of explaining it to one (or a few) person(s) here. Not saying it was bad that you just wrote a big comment explaining it completely! Just an idea of equal effort while reaching more people :)

That's very good advice, thanks for the reminder. I do edit Wikipedia fairly often, but didn't feel like I had enough time to do a creditable job on this topic this weekend. So consider my HN comment a rough draft of an upcoming Wikipedia edit.

I suspect it might have something to do with your mechanical aptitude - the mechanism is suggested by the shapes of the parts. The diagram is reminiscent of those tests we did in school to help choose a career.

What I find most frustrating about the current class of spray bottles is that they've removed the ability to unscrew the head, and therefore you can't a) get out the liquid which can't be sprayed, or b) reuse the bottle.

It's a simple mechanism, but the writing and diagrams deserve an A+ for clarity.

Wouldn't this design create a vacuum in the reservoir?

Edit: (reservoir = bottle)

There is usually a small air vent into the bottle. Small enough that fluid tension stops any spillage even when turned upside down.

Because it is small you can cause the bottle to collapse if you do a lot of spraying quickly, an effect you might have noticed before.

There is usually a vent somewhere else.

No, because the one-way valve on the bottle end is pointed the other direction, so that it allows fluid into the reservoir, but not back into the bottle.

(Too late to edit: you’re using the terms correctly, I was wrong. I was thinking of what the article calls the “staging area”)

I found his write-up on MBTA bus mirror positioning to be a short but interesting read as well: https://alexanderellis.github.io/blog/posts/mbta-bus/

Many of the buses I've seen in parts of southeast Asia have their side mirrors hanging in the front instead of jutting out the side, probably for a similar reason (and the streets there are much narrower too.)

Another reason that the right side mirror is higher is to prevent hitting passengers as the bus pulls up to a stop. Passengers tend to step up to the curb where the front door opens up as the bus approaches.

From the map I can't tell how the bus can "go through Bennett Alley and continue straight into the bus tunnel".

The streets are not aligned, are they?

Good call! You're right that "continue straight" might not be the best wording, as they aren't aligned on the map. In real life, they're also not perfectly aligned, but the entrance area to the tunnel is much wider than appears on the map, meaning they can take a slight right to continue down the tunnel. My bet is I was thinking about roads in terms of Boston roads, where going straight is often more of an abstract concept than a true direction.

Here's a streetview with Bennett Alley on the right and the bus tunnel on the left if you want to get a feel for the turn:


Unfortunately you can't see through to the tunnel from the other side of Bennett [0], but the good news is you can see the 3 buses in action!


Fantastic, thanks!

How does the nozzle diffuse?

Fluid dynamicist here who works on the breakup of larger water jets. How the jet breaks up depends on the design of the system. (I assume you mean break up, not diffuse. Diffusion is a different physical process which occurs independent of breakup.)

A low speeds (laminar flow), Rayleigh's analysis shows that a simple liquid jet breaks up due to surface tension effects.


At higher speeds (e.g., transitional or turbulent flows), the mechanisms change. For your typical spray bottle (without a swirl chamber) I'd say turbulence is the primary cause of breakup. Aerodynamic influence won't factor in at typical air densities and cavitation isn't happening as it probably would break the nozzle.

Here's a conference paper I wrote on turbulent breakup theory, which hopefully is readable for anyone with an understanding of probability theory and basic turbulence theory: https://engrxiv.org/35u7g/

(RMS velocity is just the standard deviation of the velocity. Turbulent kinetic energy is an energy formed from the RMS velocity. Kolmogorov's 5/3 theory is used a bit, but you can just view it as power law distribution. Etc.)

I also have some small spray bottles with swirl chambers, which makes the bulk fluid spin, destabilizing the interface, leading to droplet formation. The spray angle for these is large compared against the simple pressure atomizer described previously.

> Fluid dynamicist here who works on the breakup of larger water jets

One of the reasons I freaking love HN. No matter the topic, there are actual experts reading it and willing to comment.

Thank you, I was also expecting that information, not an essay on small-scale ball check valves.

I would guess it's just a fine mesh that the liquid is forced through.

It's just a hole. If you take a hot needle you can widen the hole of spray caps and increase the throughput (and size of the throw if you also increase the distance to the object).

Another trick for spraypaint bottles is to stomp them vertically if you run out of spraypaint, it won't hurt the system that takes your cap, and gives you little bit more paint.

I wouldn't think so, that's how you go from spray to foam (e.g. some household detergents around here have a small mesh screen you can clip over the spray nozzle to get a detergent foam instead of the liquid droplets of a regular spray bottle).

I think it's actually that the particles are forced to collide at relatively high speed due to the pressure behind the small opening. A mesh could eventually get clogged with minerals or debris, but this approach leaves less room for failure. Basically you're just ripping the fluid into mist with high pressure.

Having source be in the "staging area" after "git add" makes more sense to me now :-)

I think this use of "staging area" comes from the military use, which in turn comes from the theater "staging", both predating git and the bottle.

Yep, it’s more likely the author got the ”staging area” from git, as an analogy familiar to software engineers (or else from the military/theater context), it’s not an official fluid dynamics term used by spray bottle engineers. At least I don’t think so :p

Yes, I was a bit bemused by the use of "staging area" and thought the author was deliberately going for an anachronistic effect, in much the same way that (computer) machine instructions were called "orders" in the very beginning.

The more common term is just "cylinder".

I am not sure about the military use coming from the theatrical.

My understanding is that staging means 'standing place', and that the theatrical stage is just 'something to stand on'. You sometimes get scaffolding called staging in the UK.

The military use has the same root but is not derived from the theatre I beleive. A staging area is somewhere you stand and wait. Very similar in use to the staging-posts that a stage-coach goes through.

The meaning in git could mean directly this, somewhere to wait, or it could be derived from the further development of the process stages. The stage having become the journey itself rather than the stopping-point. Something is staged because you have completed a stage and are waiting to start the next stage. That is rather circular really.

At this stage I will pass you over to an article https://en.wiktionary.org/wiki/stage

If anyone's curious for more, look up how an atomizer works. It's simpler but requires more background knowledge (namely, the Bernoulli principle).

I know why I clicked on this; one of my cleaning spray bottles just broke. What's everyone else's excuse?

I was literally just thinking "how does this work" when looking at a spray bottle yesterday. Thanks!

This is cute and all, but still silly. At least, for those of us whose childhoods included dependence on wells with hand pumps. Or particularly, responsibility for maintaining said hand pumps. Because it's an ancient design. Old ones used leather flaps as one-way valves. Same for tire pumps.

The spring might be the reason why a spray bottle doesn't last very long if you re-fill it with a bleach solution, when it was not designed for it?

Great write-up, awesome diagrams, some parts could use some improvement in language, but the diagrams more than make up for it.

I would have loved this explained in my high school physics class

Clever blog! Design taking apart :) Hope he does more of these!

A clearer title would be, 'How does a spray bottle work'

I'm guessing the authors first language isn't English, but saying 'why' would be better suited to explaining why a spray bottle works for a certain solution or process.

Why does a spray bottle work well for watering a plant?

A spray bottle produces many small drops of water which is an effective method of covering the leaves of a plant.

Author here, thanks for the feedback! English is indeed my first language, but I am always looking to improve.

I chose "Why" for reasons similar to what jstanley wrote. When I was first considering the mechanism, connecting these components by simple tubes seemed like it wouldn't work. Why then did it work?

Why does a spray bottle work [when it seems like it shouldn't]?


Why does a spray bottle work [well for watering a plant]?

I saw that sctb temporarily changed the title to "How", but it looks like we're back at "Why". I apologize if any readers were expecting a more metaphysical discussion about spray bottles, but I'd be more than happy to explore that in a future post!

I thought "Why does a spray bottle work?" was a fun way to phrase it. It implies that OP originally believed, upon casual inspection, that it shouldn't work at all.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact