True HP Aero For 2011

Trichy Bastard

Well-Known Member
After considering some of the numbers to do with water hammer, I see that the pressure spikes can be 5x or more momentary pressure (5x140psi)= 700psi!!! I know it's only for fractions of a second- but will I be really wearing on my diaphram in the accumulator tank like this? What's the real world experience?

Quote from Cav's thread- which he quoted from another source that is now a dead link:


INTRODUCTION TO WATER HAMMER:

Quick closing valves, positive displacement pumps, and vertical pipe runs can create damaging pressure spikes, leading to membrane damage, pipe-work, diaphragms, seals and gaskets meters and gauges.

Liquid for all practical purposes is not compressible, any energy that is applied to it is instantly transmitted. This energy becomes dynamic in nature when a force such as quick closing valve or a pump applies velocity to the fluid.

Surge or water hammer, as it is commonly known is the result of a sudden change in liquid velocity. Water hammer usually occurs when a transfer system is quickly started, stopped or is forced to make a rapid change in direction. Any of these events can lead to catastrophic system component failure. Without question, the primary cause of water hammer in process applications is the quick closing valve, whether manual or automatic. A valve closing in 1.5 sec. or less (depending upon valve size and system conditions),
can cause an abrupt stoppage of flow. The pressure spike (acoustic wave) created at rapid valve closure can easily be as high as five (5) times the system working pressure. Unrestricted, this pressure spike or wave will rapidly accelerate to the speed of sound in liquid, which can exceed 4000 ft/sec. It is possible to estimate the pressure increase by the following formula

Simplified Water Hammer Formula: P = (0.070) (V) (L) / t + P1
Where P = Increase in pressure
P1 = Inlet Pressure
V = Flow velocity in ft/sec
t = Time in sec.(Valve closing time)
L = Upstream Pipe Length in feet

An example of pressure hammer when closing a solenoid valve, with a 50 ft long upstream pipe connection:
L = 50 ft
V = 5.0 ft / sec( recommended velocity for PVC piping design)
t = 40 ms(solenoid valve closing time is approx. 40-50 ms)
P1 = 50 psi inlet pressure
therefore, P = 0.07 x 5 x 50 / 0.040 + P1
or P = 437.5 psi + P1
Total Pressure (Spike) = 437.5 + 50 = 487.5 psi

As can be seen from the above, the pressure spike is approaching 10× the original system pressure which demonstrates the potentially catastrophic effect of water Hammer in an extreme case
 

tapar

Member
First off thanks to all you guys forging ahead and blazing a trail for the rest of us, I really appreciate it. I find the whole subject of HPA aeroponics to be very fascinating and can't wait to learn more about it. I find the mechanics behind making it work very interesting and reading about your testing and setups to very entertaining. I think I have a decent handle on why HPA is more desirable theoretically, I am still trying to wrap my head around it a bit from the practicality side. I can totally envision why getting the sexy fluffy roots is awesome and the oxygen to absorbed nutrients ratio sky high, but I am still lacking a frame of reference for how that affects things. If we were talking a tomato plant, would it grow faster or yield more tomatoes? Can other factors undo the benefit, for example is a plant with HPA but under less optimal lighting still going to show a benefit over a LPA sytem, or with the lame lighting make them more equivalent? I kind of want to get a system going just to experiment and find out, but I don't have the room to setup statistically valid tests. :)

I did have a small spark in the back of my brain go off when you mentioned looking for a way to automate adjustments to your mister timing to deal with changing environments. There is one technology that might be interesting to think about. In my own quest to try and decide if I want to join the HPA tinkering club, I wound up watching a couple of youtube videos about a leaf sensor that measure's the plants stress reactions to determine when it actually wants water, then only watering when the plant needs it to conserve water. I don't know if it would be accurate or timely enough data to really control the misting timing burst to burst, but it might be able to detect trends and allow an over time adjustment to be calculated. I am not fully convinced it is a viable solution but it is worth thinking about, here are a couple of links:

http://www.youtube.com/watch?v=LhIFntWt30c
http://www.youtube.com/watch?v=14FFgsAz5Rg
http://www.youtube.com/watch?v=skLAI2kfQiE
 

Trichy Bastard

Well-Known Member
Tapar- thanks for joining in... I am like you in that I am just trying to navigate the trails others before me have actually done the real blazing in... Anyway- I have seen the leaf sensors you mention. They are by Biocontrols/Agritech (the same place I am getting my mist nozzles from). The company seems to be involved with research and creating aero stuff for NASA among other things. I find it particularly humorous that the company is headed up by a "Richard Stoner" considering our applications here... To his credit (Under another company called GTI) is the Genesis machine and Genesis V, one of the first self contained, modular, microchip controlled aero units from back in the 1980's and it ran off mains water lines from what I can tell. Anyway, their stuff is pretty pricey overall, but that leaf sensor has been of particular interest to me. Not too much though, because I have a strange feeling that if the leaf has any signs of needing water- then the roots may already be too dry to sustain the delicate root hairs. (Although this is purely speculation on my part). I am sort of running on a weird logic here- Tree farmer is well acclimated with Biocontrols, as he recommends their misting nozzles, and he never says anything about the leaf sensors. So either it's just way too expensive to try, or his own research and knowledge has brought him to understand that it won't work the way we'd like it to. Anyway, until someone directly tells me the reasons it won't work, then it's always a possibility (Atomizer probably will explain why this can't work though). I dunno man- if you bought one and reported how it was working, I would definitely be all ears :) Perhaps it can be loosely used to indicate a "potential" for need of water at the roots, sort of how VPD indicates a potential for transpiration. The data logged over 24 hours could make tiny changes to the mist cycle and see if it corrects any abberations. Now, we get back to real life applications- where a simple recycle timer seems to be doing things well enough, and perhaps that is why people are happy doing it that way without the extra trouble. ;)

In response to your questions about the actual plant benefits in regards to the increased root efficiencies invloved in a proper HPA setup, I have no real life experience to draw from. But I have read that the plants yield more gm/watt and basically just grow faster in general with all other paramaters the same. There is also reports of tighter internodes so that the plants grow shorter and bushier. It all makes sense to me, enough to pursue trying it for myself. There are but few pics circulating, but the one's I have seen have definitely impressed me...
 

Atomizer

Well-Known Member
Hi tapar,
It looks like it works by measuring the leaf thickness so its a much better approach than using chamber RH or evaporation. I guess the question is, how to correlate a given change in leaf thickness to the ideal misting schedule. I doubt the thickness will change instantly after a misting so logically there will be a certain amount of delay or lag.. If it controlled the solenoid directly you could have 5 minutes of mist before the leaf thickness begins to change.
The best approach would be to use the sensor to control the cycle timer, setting the pause duration to match the time it takes the leaf thickness to change (initially found by trial and error and assuming its a constant) and use the shortest pulse possible. That way the sensor will only trigger a single mist pulse and then wait long enough for a change in leaf thickness to occur before deciding if another pulse is needed to reach the target thickness.
You`d be in for some extensive field testing and logging to figure out how leaf thickness changes relate to maintaining the ideal aero root structure but you may be lucky enough to find a single leaf thickness differential and short mist pulse is enough to automatically maintain the ideal mist environment throughout the entire grow...aka holy grail ;) negating the need to constantly alter the mist parameters to account for plant size, growth stage and environmental factors. One setting may be all you need, if it exists and if you could discover it from testing ;)
A lot of question marks but it looks more promising than any other approach to mist automation.
Anyone rich enough to buy one of these is likely to have plenty of free time to do the testing..any takers? :)

Not as easy as we hoped:
http://jxb.oxfordjournals.org/content/43/3/327.short
 

Trichy Bastard

Well-Known Member
I always get excited by the abstracts from these journals, and get bummed out that I can't read the rest without a paid subscription... Anyway- I guess the best approach might be to first decide how a human can detect the needs, and try to automate that process. In this case we look at the roots and decide if they look wet or dry enough and if they've been getting the fuzzies. So I guess the best approach would have to be a very complex optical recognition system where it can interpret what a camera in the chamber sees, and make calculations and changes to the mist timings from there. As far as I know NASA isn't even doing that- so guess the holy grail remains just that- something unattainable without a major revolution in technological advancement . But it all starts with an idea, and we should never stop thinking. I remember reading a couple thousand years ago that people somewhere decided to stop doing scientific research, the king claimed that everything that could be invented already had been done. Obviously they hadn't any access to the internet to realize they were wrong... ;)
 

tapar

Member
Ouch, at $290 for snagging just the sensor, that is a swift kick to the pocketbook! I wish my wallet wasn't so wilted, it would be fun to play with one! I was curious what the price was and while trying to navigate the AgriHouse site I managed to find a few interesting tidbits. It looks like their software measures the turgidity (water thickness) of the leaves and I think it uses an over-time average to determine what 'not thirsty' is, then when they detect the first signs of that thickness dropping and when it crosses a threshold they consider that 'thirsty'. They actually use this system with their aeroponics unit, here is a link about that: http://leafsensor.wordpress.com/2010/09/24/leaf-sensors-for-true-aeroponic-growing/

It looks like they do try to use the leaf response time to control when to send the mist. The link Atomizer provided highlighted the difficulty of calibrating this kind of data to a particular plant in a particular growth cycle. The paper referenced was from 1991 so it is possible that advances in automated measurements have made it more practical now. If they were measuring by hand it would probably be difficult to monitor the readings in real time. If they are getting a continuous feed of data it is probably easier to detect changes that are not related to aging or growth spurts by using a moving average of the thickness. It would still be specific to that plant but at least the data would be more meaningful.

I don't think the AgriHouse aeroponics units sound as fine tuned and dialed in timing wise as the ones you guys are working on. I would guess that (presumed) lack of precision might allow more room for something like this to not mess everything up in one of their systems.

I think experimenting with this could be really fun if it were not so expensive to get a test unit. I always thought anything that automates a system is pretty awesome. I think Trichy Bastard is right and it is not practical enough to experiment with. I would also think you would have to dial the system in initially to generate fuzzy roots in the first place. If you don't have fuzzy roots...I would expect the sensor to demand water till you had some...but if your over-misting...you won't grow them. It sounds like a chicken and egg kind of situation. You probably could field test your way to good enough initial settings and then use the sensor from there out. I don't have a good grasp of how much work goes into keeping the fuzzy roots after your initial dialing in of the system, or how strain specific that info is.

This seems to me like something you might play with after you have a system up and running for a while and all dialed in and looking to invest in tweaks to the setup. :)
 

Atomizer

Well-Known Member
So I guess the best approach would have to be a very complex optical recognition system where it can interpret what a camera in the chamber sees, and make calculations and changes to the mist timings from there
A seasoned grower with half decent eyesight :)
 

Trichy Bastard

Well-Known Member
I don't think the AgriHouse aeroponics units sound as fine tuned and dialed in timing wise as the ones you guys are working on. I would guess that (presumed) lack of precision might allow more room for something like this to not mess everything up in one of their systems.
Bingo! Your looking at a $5000+ setup there with their Genesis units- and they are no where near as advanced as what were going to do here. I would not expect their roots to take on the truly efficient fuzzy structure given the 3 second spray times, etc... So when your girlfriend/wife complains how much your throwing at your new hobby- tell her your actually saving money AND making a better system... ;)

To be honest, being that they play heavily on their ties with Nasa, I was quite surprised to find out that that they were not nearly as bleeding edge as it would seem. Their leaf sensor will work fine for their half-ass aero system, but ours is so much more fine tuned, the sensor will probably serve only to drag the results down to what theirs are... The other irony is it talks about how the sensor will save ya 20% on watering and nutes, but IT'S A RECIRCULATING SYSTEM TO BOOT! I don't know about you, but the fact that we can DTW more economically than regular hydro costs to recirculate, nevermind electrical costs, is probably the best selling point of this system overall... I'm feelin pretty good about my choice in system right now...
 

rosecitypapa

Active Member
RCP, You could in theory time down to the micosecond. The average 328 ATmega instruction takes 1 to 2 clocks to complete, at 2 clocks per instruction(@16mhz) we are in the range of 8 instructions per uS. All you would have to do is fit the timer cycle into 8 instructions for 1uS resolution. Of course none of this really matters since we are interfacing electromechanical equipment with much higher delay characteristics, so we find our highest delay part and that is our "switching" time. In my case that will either be the relay or the solenoid(@20ms), I have both SSR(@2ms) and MR(@40ms) so I will see how they both perform. I really dont see why you can achieve what you want, I have most of it running right now just need to finish off the HPA side now.
Kdn, what is SSR and MR?


It looks like it works by measuring the leaf thickness so its a much better approach than using chamber RH or evaporation.
Atomizer, why wouldn't measuring root chamber RH be an effective way to automate timing pulses?


Ouch, at $290 for snagging just the sensor, that is a swift kick to the pocketbook! I wish my wallet wasn't so wilted, it would be fun to play with one! I was curious what the price was and while trying to navigate the AgriHouse site I managed to find a few interesting tidbits. It looks like their software measures the turgidity (water thickness) of the leaves and I think it uses an over-time average to determine what 'not thirsty' is, then when they detect the first signs of that thickness dropping and when it crosses a threshold they consider that 'thirsty'. They actually use this system with their aeroponics unit, here is a link about that: http://leafsensor.wordpress.com/2010/09/24/leaf-sensors-for-true-aeroponic-growing/
That would be a fun diy. Low cost biofeedback equipment based on galvanic skin response linked with an Arduino would be a good start. Ebay has them for $25.
 

Kdn

Member
RCP, short for Solid State Relay, Mechanical Relay. (Upon reviewing my MR datasheet its actually @28ms)
 

PetFlora

Well-Known Member
At some point you have to go analog (eye balls). The more people rely on devices, when issues develop (in HPA you can count on it), the less likely they will know how to respond.
 

Atomizer

Well-Known Member
Atomizer, why wouldn't measuring root chamber RH be an effective way to automate timing pulses?
I`ve answered this one a few times in other threads but it keeps cropping up :)

RH is a measurement of the water vapour/air mixture, water vapour is a gas which is unavailable to roots until it condenses back into liquid water. If you fire a short mist pulse into a chamber and time the rise of the RH you`ll find it can take 3-5 minutes to reach its peak.
If you time how long it takes to fall back until it stabilises..it will be a longer (10minutes or more) and it may not fall to the original start point.
Using trial and error, i found a cycle time of 0.7sec/ 137sec produces good results during early growth, so its impossible for me to use RH to trigger the mist because the chamber RH wouldnt even reach its peak value in the time available before the next mist pulse.
My outdoor chamber has a remote probe installed which is attached to a min/max hygrometer/thermometer, the min/max RH is showing an 8% swing since i reset it about 2 weeks ago.
 

Atomizer

Well-Known Member
At some point you have to go analog (eye balls). The more people rely on devices, when issues develop (in HPA you can count on it), the less likely they will know how to respond.
You`re into the trial and error either way. You`d need all the answers in order to calibrate your automated control system :)
 

hammer21

Well-Known Member
rb30rl-2.gifGoing to be picking up a dozen of these 30 gallon drums this coming week used and clean with rubber gasket on the lid 10 buxs a peice . Will be going with 3 nozzles per drum facing down with 1/2" bulk head fitting in the bottom for a drain will be using a type of silk screen on the bottom to prevent roots getting in the drain. $(KGrHqR,!lQE3HG8S)H4BOCmfHpsh!~~0_12.jpgI will be using 24 2-way solenoid valves and 12 3-way solenoid valves going to try using this type of fitting to blend my nutes and pure oxygen threw the nozzles. Also will be able to purge the whole system to drain before changing nute mix. Already got the tube reducers to fine tune the mixture. Will be using standard industrial oxygen for cutting torch`s have it. Thats whats going on with me. Also just about done with a new controller build will post some pics when completed.
 

Trichy Bastard

Well-Known Member
View attachment 1677710Going to be picking up a dozen of these 30 gallon drums this coming week used and clean with rubber gasket on the lid 10 buxs a peice . Will be going with 3 nozzles per drum facing down with 1/2" bulk head fitting in the bottom for a drain will be using a type of silk screen on the bottom to prevent roots getting in the drain. View attachment 1677721I will be using 24 2-way solenoid valves and 12 3-way solenoid valves going to try using this type of fitting to blend my nutes and pure oxygen threw the nozzles. Also will be able to purge the whole system to drain before changing nute mix. Already got the tube reducers to fine tune the mixture. Will be using standard industrial oxygen for cutting torch`s have it. Thats whats going on with me. Also just about done with a new controller build will post some pics when completed.
Cool Hammer... I'm not sure adding o2 will be beneficial. The atmosphere is around 21% oxygen and in high concentrations it can become toxic to humans (not sure about plant roots). I would assume adding a small air hose to dribble in some ambient air if the chamber is completely airtight would be plenty- but that's purely conjecture on my part. Of course messing with high levels of o2 can be dangerous because of the ease which things burst into flames, but I'm sure you're already well aware. Gotta wonder how much o2 roots really need considering they're usually buried in the soil. Hopefully someone with more knowledge will chime in with their perspective. I like those chambers though- I'll assume your situation negates the need for extra root chamber insulation- that certainly keeps things simpler... Also after looking around in Cavadge's thread, Blick art supplies has "140 tpi monofilament screen", it has been recommended to work well as it's small enough to keep in roots, but permeable enough to let most water drain. We'll need to make sure the screen is sealed tighly in, otheriwse the roots will push their way through any minute opportunities.

I'm glad your getting on with your build- I'm still waiting for stuff to arrive and wish I had more to write about... Once it all gets here though, things should come together pretty quick and I should be back on topic writing up a storm with pics.
 

kmbud

Member
Thanks PF, I am still trying to visualize how the plants stay in the pvc adaptor before you add the neoprene, but Atomizer tells me I should try 2" net pots with a layer of rockwool in the bottom, set the seed on one layer of the cubes, and then lightly fill over it to the top with more rockwool chunks without packing down. Apparently the roots will fill in the spaces and lock it all together- sounds pretty good to me. I'll probably put an inch of hydroton over the top once the sprouts are 6" tall or so- but just for aesthetics. Cubed rockwool looks like fiberglass insulation to me- and I'm a perfectionist... I'm really excited- I just ordered most of my components today. I will update the second post in this thread to reflect what I bought and where with prices in the next day or so... BTW I found a smokin deal on a 150 psi flojet pump in the most unlikely of places, they have 6 more on hand for only 75$/ea- it's a steal- so anyone interested give a shout.
Hey TB I'm shouting about the pump if it's not too late.
 

kmbud

Member
Also I would like to know what brand and model number if you have it of the pressure relief valve and pressure regulator, you say both are SS and I can't seem to find anything close to that.
 

Trichy Bastard

Well-Known Member
The prv I found on ebay- just search for stainless pressure relief valve- 12 bucks... Hope they're still around- although it was not the best prv in the world I hope I'll never need it to work anyway ;) Also read treefarmer was using a brass one, with a checkvalve just before it to prevent the copper from getting into the main solution flow... The regulator was made by watts (model p60) found one here among other places- I got it from these guys because they had the best JG fitting prices I could find at the time too: http://h2odistributors.com/ will pm you the pump... It was days of work finding this stuff, not an easy task I know...

I caught something in Cav's thread I hadn't noticed before. This flojet pump I got is the same as his. He did say although the Aquatech pumps moved a bit less fluid, it was much quieter...
 

hammer21

Well-Known Member
Hello Trichy you are right to much of a good thing is deadly!!!!!!!!!! I plan on using a small amout of pure oxygen in the mix and will get a oxygen monitor for testing pod and room NEVER try this without the right equipment and knowledge on what will and can happen.
 
Top