Starting with Research
We had to start by looking at how people are diving to know where to begin. Divers get their air supply from their scuba tanks, which lets compressed air into a scuba regulator, that delivers air at the exact pressure they need to create the exact amount of flow of air they need.
Scuba regulators come in two stages. Stage one steps down the extremely pressurized incoming air from their tanks to an intermediate pressure, which is dependent on the ambient water pressure, usually 9 bar above it. Stage two comprises of a demand valve, a mechanism that only allows the air from the end of the first stage at intermediate pressure to reach the diver’s mouth when the diver breaths in.
It can be observed here that there is a balance between the spring and ambient water pressure against the intermediate air pressure. This is offset when the diver draws his breath through the events in stage 2, the intermediate pressure drops and causes the piston to open the chamber to the very high pressure from the air tank, which rushes to the intermediate pressure chamber, increasing the intermediate pressure just until it is enough to move the piston back, allowing the cycle to continue. It is this balance that steps down the pressure from the tank.
We observed that the second stage regulator makes use of a diaphragm and lever mechanism, the diaphragm contracting due to the pressure drop in the second stage chamber when the diver draws his breath, causing the lever to move and allow air from the intermediate chamber into the second stage chamber, increasing the pressure in the second stage just enough to balance out the ambient water pressure and push out the diaphragm, resulting in the chamber returning to ambient water pressure. This will step down air from whatever the intermediate pressure is to ambient water pressure, so long as the intermediate pressure is higher.
Development of Working Mechanism
We need to ask: How would not having a First Stage Regulator (FSR) affect the working mechanism?
Lets first look at a scuba regulator.
The main concept to focus on is that at the end of the FSR, the air output should be at a pressure within a certain goldilocks range relative to the ambient water pressure, not too high and not too low.
Now what happens if the intermediate pressure entering the second stage is too high (>>AWP)?
Firstly we make some assumptions, that the intermediate pressure is obscenely high, that the demand valve mechanism does not immediately give way and is forced open by the high pressure outside and that the diaphragm moves freely.
The diver breathes in as per usual, compressing the diaphragm and moving the demand valve normally before he has finished inhaling. Due to the pressure difference, the high pressure air from the intake rushes in with a high flow rate, much higher than the diver’s instantaneous inspiratory flow rate can remove. Air pressure in the second stage rises quickly due to this difference, however is abruptly halted when it slightly exceeds the AWP and pushes the diaphragm out to close the intake. Since the demand valve opened and closed so quickly, the diver has yet to finish fully taking in his breath! The demand valve would have to continue this cycle multiple times before the diver inhales even once, which would be quite exhausting for the diver having to use his breath to compress the padding on the mask so frequently.
Now what if the intermediate pressure is too low?
If the intermediate pressure is equal to or lower than the AWP, as you can imagine, air that is supposed to enter the second stage regulator will not even enter, due to the lack of positive pressure difference.
That does not mean any intermediate pressure higher than AWP will do, however. The intermediate pressure has to be high enough to create a large enough pressure difference to result in a flow rate similar to the inspiratory flow rate the diver finds comfortable. Less than that, and the diver will find themselves trying to suck air from the compressor! In fact, typical intermediate pressures are set by the FSR to be around 9bar above AWP. This seems like a lot, but considering that deep divers would be going under for hours at a time traversing the deep blue, a higher pressure difference and higher flow could supplement the effort they need to inhale when they are tired. In addition, pressure difference is not the only factor. When determining flow between two points, the airway diameter also determines flow, looking at the Hagen-Poiseuille Equation. A small airway may be another reason commercial second stage regulators can get away using intermediate pressures of 9bar above AWP.
Well how would this translate to our product?
This gives us the conditions we need to satisfy in order to not require a FSR, allowing us to focus on developing the second stage regulator.
The initial design would be one that outputs a single constant pressure, where we would be making use of a compressor, pressure switch, an air tank, pressure gauge and a demand valve to mimic a second stage regulator. (Model can be found below in the “plan so far”)
This design would work on the idea of reserving air, having a compressor (Maximum pressure >5bar, Free Air Delivery, FAD >90L/min), connected to an air tank (Free Air Delivery in this case being the flow of air out the compressor at the pressure of the air through the intake, 1bar). The air tank would be connected to a pressure switch rigged to the compressor, air tank and battery such that it keeps the air tank between an upper limit (set to the maximum air pressure output of the compressor) and a lower limit set at at least higher than the lowest pressure of the “goldilocks range” as mentioned above which would be the pressure that allows a comfortable flow of air to the diver at the deepest depth of 3m. The pressure switch would be connected in series with the compressor and battery, completing the circuit when the air tank hits the lower limit and opening it when the tank hits the upper limit.
The air tank would let air into the air hose leading to the diver through a filter and a pressure regulator (controls the pressure output from the tank), set at the pressure switch’s lower limit. This meant that the air tank is constantly outputting one fixed pressure, with variable flow into the mask (second stage regulator) as AWP changes.
The lower the maximum depth the diver will be diving, the higher the lower limit of the pressure of this goldilocks range will be, due to 2 reasons. Firstly, we would actually require more air the deeper we go, about 1.3 times as much as we would on the surface (as the flow, L/min we require is constant, a higher pressure with the same flow would mean more air required, more on this in learning outcomes), hence a higher pressure difference to create this higher flow. Secondly, as mentioned, AWP rises, eventually exceeding the upper limit that one would be comfortable with on land, which is when you will start to require a first stage regulator. This should not be an issue however, as much of the examples of people doing SNUBA we have found have no problem with just a second stage regulator breathing on land with their set-up meant to dive much deeper than 3m.
Designs for a second stage regulator/demand valve:
We initially had ideas to use the rubberized nasal section that certain snorkeling masks come with as a diaphragm. This would be connected to a hatch attached to the snorkeling air tube on the mask via a string/plastic strip, pulling open the hatch as the diaphragm contracts and extending as it relaxes. The hatch would have to be almost air tight (if its not excess air would just come out the exhaust), and normally closed (spring back into closed position).
More ideas included transplanting the mechanism of an existing store bought second stage regulator onto the mask, or implanting a custom made diaphragm into the mask during the brainstorming phase. Eventually we realized the flaws of such methods as well as our own manufacturing limitations (especially for a breathing mechanism where durability and reliability is crucial) , and were stuck on how to proceed. Then one day, we had an epiphany.
Creation of the Demand Valve
The first and probably most important realization was that many typical full face snorkel masks have characteristics that may be adapted to mimic the features of the second stage regulator, mainly the elastic compressible rubbery padding around the mask and the exhaust valve.
Padding
The padding on the plain mask serves merely as a watertight seal between the mask and the diver’s face.
As there was only one main force acting on the mask and its padding that changes (ignoring additional drag from the water while swimming etc), the ambient water pressure acting on the outside of the mask, (forces such as the straps attached to the mask pulling it into the diver’s face, the normal reaction from the diver’s face and gravity are not expected to change), the compression of the padding would solely be determined by the balance of the ambient water pressure and the elastic force of the padding.
We realized that the padding could somehow be linked to a mechanism that would trigger the airflow from a compressor on the surface, compressing as the diver inhales and expanding as the compressed air flows in to increase the pressure in the mask.
Exhaust Valve
Many full face snorkeling masks come with an exhaust valve where air exhaled by the diver exits out, into the surrounding waters, while also preventing water from entering.
Rubberized nasal section:
Looking at diving physiology, you see that our middle ears are dead air spaces, connected to the outer world by the eustachian tubes running to the back of your throat. Failing to increase the pressure in your middle ears to match the building pressure around you, results in harm to the delicate mechanisms of the ear, and lots of pain.
The key to safe equalizing is opening the normally closed eustachian tubes, allowing higher-pressure air from your throat to enter your middle ears. Most divers are taught to equalize by pinching their nose and blowing gently. Called the Valsalva Maneuver, it essentially forces the tubes open with air pressure, and is a common practice with almost all divers. Conventional snorkeling however, does not normally pose such a concern, as a result most full-faced snorkeling masks do not come with such a feature, with the exception of a specialized few that are meant for allowing the user to make short free dives while holding their breath.
Airflow:
A common design of these masks is to have the lower section of the mask around the diver’s nose and mouth to be secluded from the rest of the face, separated by a one way air valve that lets air in from the upper portion.
Inhaling:
Initially the air pressure in the lower and upper sections are equal at 1bar.
The diver inhales, pressure dropping in the lower section of the mask, air enters from the snorkeling tube, to the upper section, through the one way valve, to the lower section due to the pressure difference, where it is inhaled by the diver. The one way valve closes as the lower and upper sections are again equal in pressure.
Exhaling:
The diver exhales into the lower section. Air pressure in the lower section momentarily increases. As air cannot exit through the same one way valve by which it came, it is forced through two paths, partially through the exhaust valve at the bottom of the mask, and partially through a separate airway leading out upwards to the surface, alongside the intake tube, which also contains a one way air valve only allowing air out. The resulting air pressure in the mask is at atmospheric pressure.
The ‘Regulator’
The next realization would have to be for the simple and elegant solution for the mechanism used to link the new diaphragm and the hatch.
A lever switch was the ideal triggering mechanism for the compression of the mask padding. The spring in the switch ensures it would stay normally open, it was small, compact, and durable, it even allowed the user to have a feeling of the switch’s mechanism clicking on and off while placed in the mask. Not all lever switches would perfectly fit our intentions however. We kept this in mind when we ordered all kinds of designs of lever switches to see which would work the best.
Some things to keep in consideration would be the elastic force of the spring when contracted for the switch to be open (too much could cause discomfort over time), the size of the switch (all of them were generally small but some were bulkier than others, some of the dimensions may not have been included in the description) and if it were waterproof or not.
*The length the switch has to be pushed and the length the padding contracts when the diver breathes in were not as key as the switch could just be adjusted to any position to accommodate how much effort one would need to exert when inhaling to compress the padding to trigger the switch. We measured the average length each of our members could contract the padding by at a comfortable exertion to be an average of 3mm.
As it would be electronically operated, the hatch would now be a solenoid valve. As it would need to open when the padding is compressed and switch is closed, the valve would have to be normally closed.
Here we have the switch attached to a multimeter, beeping whenever the circuit is completed, allowing us to test quite a few things concurrently.
-Which of the switches that we purchased would be ideal?
-Would this mask compress sufficiently to trigger the switch?
-Which position would allow the switch to trigger when exerting a comfortable enough inhaling pressure? (the switch should not be positioned too close to the edge of the mask padding such that it requires so little effort to trigger, any slight disturbance, such as the hose tugging on the mask, would activate the compressor)
The Plan so far:
The compressor draws air from an upright tube at the surface for air intake, outputting compressed air into an air tank through a one-way valve. The air tank would then be attached to a pressure regulator (a valve that allows the adjustment of what pressure passes through), to an air hose leading down to the air tube of the mask.
Small Battery powered (solenoid valve, small battery and microswitch)
Main Battery powered
The demand valve could either be powered by an extra small battery, or be connected to the battery with wires leading up to the battery along the air hose. The connection of the snorkeling air tube to the hose was also another hurdle, as the two pieces were very different in size, with the tube being much larger. After scouring for the right snorkeling mask and purchasing it, we saw that ours came with a detachable attachment (a mechanism that keeps water from entering when submerged, as this was a snorkeling mask made also for short dives).
We 3D-printed a custom piece to link the tube to the hose, with one side in the shape of the connecting piece to the snorkelling tube, and the other with a hole the size of a 1/4 inch NPT screw attachment for the solenoid valve.
Recreating the demand valve:
This ‘constant pressure’ design meant for a heavier compressor that would be able to output such a high pressure, as well as a larger and bulkier product with the inclusion of an air tank. We would also come to find an unexpected limitation of the compressor when we actually tried this design as well as potential risks, (further explanation of this limitation can be found in the learning outcomes page), that forced us to reconsider the entire air-supplying mechanism.
To do away with the air reserving design, we had to leverage a key property of certain compressors. Positive-displacement compressors output constant flow. With this in mind, we realized this was much more straight-forward, as regardless of pressure increasing with depth, people consume the same flow of air (more of this in learning outcomes). This way, we could just focus on the compressor outputting this constant flow, allowing us to cut out the constant pressure design, along with the solenoid valve, air tank and pressure switch.
Now that we would have to settle on a singular constant flow, we needed to know, how much?
The average air flow demanded by an individual varies greatly depending on the level of exercise, from 6L/min at rest to 90L/min while undergoing heavy exercise. Lung volumes and vital capacity – Cardio-respiratory system – Eduqas – GCSE Physical Education Revision – Eduqas – BBC Bitesize. Every individual would of course also have their ideal own upper and lower limit for the flow into the second stage, depending on how quickly and slowly they can naturally inhale comfortably. This changes with their metabolism, emotional state, and every other invisible factor that pushes their own fickle preferences on how much air they want to breath at any point in time.
This average consumption of air would be the equivalent of what Scuba Divers call their SAC rate, which is an estimate of how much flow an individual should be consuming in each dive, depending on how much they plan for the pressure to drop in their tank, the volume of their tank, and the average depth they would stay at.
In our case, it should be much simpler, assuming the diver on average stays close to the maximum depth we intended, 3m, the air ending up in their lungs would ideally be at 1.3bar, so our compressor just needs to have a flow of 90L/min at 1.3bar right?
Answer: Well no, but actually yes.
Firstly, keep in mind that this is an estimate for the average airflow across multiple breaths, including the time when the diver exhales. This is known as the minute ventilation. However, the demand valve/compressor would only on and provide air flow whenever they… demand it, when inhaling, of course.
Thus compressor flow= (Minute Ventilation)/ (percentage of time spend inhaling)
Secondly, we wouldn’t expect the diving experience to be a constant, full body heavy exercise. For even an amateur scuba diver, lugging around their air tank, breathing unfamiliarly through just their mouths, their minute ventilation would just be 25-27 L/min. What is SAC Rate, How To Calculate And Manage It? – Scuba Diving UK (surfaceinterval.uk)
Microsoft Word – Document1 (weebly.com) This article suggests an important rule being that the time spent exhaling should be twice as long as the time spent inhaling for efficient swimming. While we expect the diver using our product to have a much different experience (even when trying to swim for distance, the diver may be using flippers, swimming completely submerged and pulling along the product floating on the surface) from that of swimming in a pool, this ratio felt natural for us. We would do a mini trial, measuring the average length of time each of us spends inhaling on land, trying our best to mimic the breathing patterns one would have when scuba diving, and find that the ratio of time spent inhaling to exhaling was close to 1:2.
Working with these assumptions and averages, the compressor FAD flow we would need would come to 75-81 L/min.
We would start by trying a compressor with a max working pressure of 1.85bar and flow of 65L/min at that pressure, which would be a sufficient 92.5L/min at 1.3bar.
Now the new design for the demand valve would still use the same microswitch- padding mechanism, only that it would on and off the compressor with each breath.
After some suggestions taken from the lab advisors, we also decided to add more components to the circuitry, mainly the buck and relay, to reduce power going to the mask. (further elaborated in learning outcomes)
Pilot test:
The product is to be assembled on site. Firstly, connect the hose to the filter, and the wires leading from the mask to the main circuitry. Then turn on the larger switches and press on the microswitch to see if it triggers the compressor appropriately. Fit in the floats and wear the mask. Try breathing with the device on land to see if the microswitch triggers comfortably. If not, the position of the microswitch would then need to be adjusted. This would be expected to differ from user to user, however we found that at least for our teammates it did not need to be readjusted. Once this is done, dive underwater, leaving the compressor on land. If everything is good then you’re ready to go.
We would realise during testing that the average time the compressor was on was 32%, with the average ‘burst’ of the compressor ranging from 0.7s to 2.0s. This proves the estimate used earlier to be rather accurate. We cannot make any estimate for our average airflow inhaled, however, as any excess would have come out the exhaust. The pool test would clear many of our smaller worries: no significant force was felt tugging the diver from the hose, the diver is able to ascend and descend without the tube getting in the way, and air was reliably supplied by the demand valve and compressor (with one little hiccup, more on that on learning outcomes). The concern of stability and risk of capsizing was quelled as well, with the floats in place around the product, the entire float was extremely stable in water, allowing the diver to rest their weight on it in the water without worry of sinking or capsizing. Hence for our intents and purposes, we would not expect the device to capsize due to any strong waves. A long upright tube to prevent water from accidentally splashing into the intake of the compressor would later be included.
Other tedious hassles leading up to the test and on the pool day itself included planning and assembling connectors to our various components, waterproofing the various components in contact with the water, and filming underwater.
Further Developments
Communication
The future of the SASS is very bright, with the unique design of the demand valve opening up many promising prospects.
One such prospect would be the addition of a microphone and bone conduction headphones inside the mask. Bone conduction headphones have to be used to allow the user to hear underwater with their ears submerged. Both this and the microphone would be connected to the users smartphone secured on the float above, allowing the diver to make calls and communicate with each other underwater.
Even advanced specialized scuba helmets on the market have to rely on bluetooth, severely limiting their range of communication to nearby divers with the same scuba helmet device (bluetooth does not travel very well in water at all). Our product would have no such limitations, can would be able to communicate with people on land and fellow SASS users anywhere.
It is only because of the fact that our product has both a tether to the surface AND keeps the user’s mouth free that this feature may be implemented.
Deeper Depths
Due to access restrictions, we could only try the product in a 4m diving pool, and as such dive into deeper waters to explore the limits of the SASS. In an optimistic future, we would like to extend our product to depths of 10-15m. This would mean modifications including the addition of a FSR, and reverting back to the ‘constant-pressure’ model with a small air tank (large enough such that the pressure drop with each breath will not cause the pressure of the tank to drop lower than the intermediate pressure), pressure switch and pressure regulator.
Battery Gauge
