crazzyreefer
New member
fixed a few glitches this should work better
fixed a few glitches this should work better
Version 1.1
fixed a few glitches this should work better
Version 1.1
Skimming Principles
- Thread starter tinygiants
- Start date
crazzyreefer
New member
Dammon1 the skimmer in the drawing needs to have 13 feet of tubes and a 10 gph pump, running counter current, and a air pump that delivers 0.5 cfm. through a vary fine diffuser
This is a vary workable model!
This is a vary workable model!
tinygiants
New member
I will admit that the math in the formulas is beyond my skills, but I believe that the rate of rise is 28 to 30 cm/s. Did I miss a conversion needed for our bubbles?
As I played with the numbers some things did not seem right. When I increased the flow rate to 120 gallons the cocurrent dwell time was still 134 ( nothing else changed). In a cocurrent design, that would be pushing the bubbles to the surface. That number seemed high in a cocurrent set up. The counter current number is only 8 seconds longer.
Is that the right math?
Thanks,
Dale
crazzyreefer
New member
you stated 4.5, not sure if that was inside or outside dia. but its of no consequence to dwell time. due to bubbles travel up, and tube size only affects your air to water ratio.... but I re-looked at your skimmer with a mod, I can reduce the tube to a few feet...or inches! hear me out... take a skimmer, at the base use a large pump as a vacuum, a variable speed motor would help....a few inches above it, place the air stone, the bubbles will rise, so adjust the pump to suck the water out from the bottom almost as fast, but not too fast to suck in the bubbles... the bubbles will be on a treadmill of sorts until they reach the top, you can literally shorten the tube to only a few inches, and still have dwell time of 120+ seconds and have nearly a saturation point of 12-13%, and the skimmer can be made in any size including HOB, It dawned on me as I was trying to figure out how to make a more practical and affordable model... then I thought about convection, like an oven.... so I temporally named this a convection skimmer... I know convection is only used in heat... well anyone have a better name?
crazzyreefer
New member
yes that the right math... its take from a proven formula, buble rise is 1.17 inches per second in still water, the bubble will rise 1.17X8 = 9.36 inches further or you will need 9.36 more skimmer body.
tinygiants
New member
I agree with the idea of moving water causing the bubble to slow down. That is the whole reason that counter current is more efficient. But we need to remember the other principals in a design.
Turnover rate - The goal is 2 - 3 times a day
This number will dictate your skimmer feed rate.
Skimmer size and design (convection, co current, counter current.....) now factors in to overall bubble dwell time.
Skimmer size is directly related to water dwell time.
Bombardment rate = water dwell in seconds / air dwell in seconds
According to Escobal the goal is 10 in this ratio (Is this goal reasonable? Is it still a valid goal?)
So while I agree that water movement can be used to lengthen the bubble dwell time it will directly effect other aspects of the skimming environment.
Turnover rate - The goal is 2 - 3 times a day
This number will dictate your skimmer feed rate.
Skimmer size and design (convection, co current, counter current.....) now factors in to overall bubble dwell time.
Skimmer size is directly related to water dwell time.
Bombardment rate = water dwell in seconds / air dwell in seconds
According to Escobal the goal is 10 in this ratio (Is this goal reasonable? Is it still a valid goal?)
So while I agree that water movement can be used to lengthen the bubble dwell time it will directly effect other aspects of the skimming environment.
tinygiants
New member
The bubble dwell time of 120 seconds causes a bombardment issue.
Using the formula above to get a ratio of 10 the water needs a dwell time of 1200 seconds.
This means the water needs to sit for 20 minutes in the skimmer body. In my system I need to feed the skimmer 2 gpm for proper turnover. So I need a 40g skimmer body to get the ratio right.
Using the 1.17 inches/second of bubble rise I need a rise travel distance of 11.7 feet. Now I need to factor in what shape gets me 40gallons at 11.7 feet travel distance.
I love how everything drives everything in this conceptual process.
Keep up the innovative work.
PS. My skimmer pump just died, so I am going to be getting serious about this.
Using the formula above to get a ratio of 10 the water needs a dwell time of 1200 seconds.
This means the water needs to sit for 20 minutes in the skimmer body. In my system I need to feed the skimmer 2 gpm for proper turnover. So I need a 40g skimmer body to get the ratio right.
Using the 1.17 inches/second of bubble rise I need a rise travel distance of 11.7 feet. Now I need to factor in what shape gets me 40gallons at 11.7 feet travel distance.
I love how everything drives everything in this conceptual process.
Keep up the innovative work.
PS. My skimmer pump just died, so I am going to be getting serious about this.
crazzyreefer
New member
tinygiants your correct, but the ratio number not to exceed is 13 not 10, salinity will also affect ratio but in out case that's not important. three items of unknown exist in the equation, first, bubble dwell is calculated as clean walled or absent of organics and oils, as a bubble dwells longer it attracts an unknown amount of organics causing drag, and allowing the bubble to dwell for a longer period or reversing direction if in a counter current method is applied. also no known data of ratio vs tube size exist..yea ill work on it, basically a 1/2 hose we know if we produce 100 gph through it will force the air to travel with the current, but a tube of 6 inches 100 gph will only slow it down,so your partially right about this but skimmer body size isnt that important, due to the fact air is moving upward, in a ocean a bubble will rise at the same rate as in your nano tank. another factor to be reckoned with you will have to know what dissolved oxygen you are feeding the skimmer, remember your tank water isn't void of oxygen, ill work on this it will affect the cfm number.
crazzyreefer
New member
120 second = 2 min not 20 min.... the average skimer is less than 20 seconds and thats the good ones! the ratio is air water mixure CFM vs GPH increase gph and your ratio lowers, but not your dwell time, sort of... dwell time is affected due to the tread mill effect.....so boost CFM!... love it when you change just one little thing and the numbers goes in an opposite tangent!
crazzyreefer
New member
convection skimming
http://reefcentral.com/gallery/showphoto.php?photo=142146
http://reefcentral.com/gallery/showphoto.php?photo=142146
crazzyreefer
New member
and if its not complicated enough, DO will affect the CFM, but vary few of us has a do tester, and ORP and REDOX will affect the dwell time due to the organics, so even though your ratios are perfect, they will change with PH, ORP,DO and REDOX this is where making varible speed pumps and air pumps with valves will fine tune the system.
ChemE
New member
Glad to see my favorite thread is still alive and kicking. With regard to the size of tube affecting bubble dwell time the conversation has once again moved into my circle of knowledge. Since this is the advanced topics thread I'll go ahead and not pull any punches; I don't feel the need to pander here.
Fluid flow velocity is a factor of 4 parameters: fluid viscosity {μ}, fluid density {ρ}, fluid velocity {v}, pipe diameter {D}. This neglects skin friction losses which in skimmers is trivial. These four parameters are used to calculate a dimensionless ratio or number called the Reynold's Number. It is dimensionless because all the units cancel each other out leaving only a ratio. This ratio is the ratio between the viscous forces and kinetic forces in a fluid flow situation. When viscous forces dominate flow occurs in streamlines which don't intermingle (think good drivers on a very packed interstate where no one is changing lanes). When kinetic forces dominate viscous forces, the streamlines start to swirl and intermingle (think 1,000 car pileup happening at 100 mph or if you've ever driven in downtown Rome...).
The Reynolds number is given by Dvρ/μ. Where the numerator is the magnitude of the kinetic forces and the denominator is the magnitude of the viscous forces. Reynold's numbers below 2,000 make up the laminar flow regime. Between 2,000 and 2,500 is a transitional flow regime. Above 2,500 is the turbulent flow regime.
In the laminar flow regime, fluid velocity changes significantly between the wall and the center of flow. At the wall the flow velocity is zero. This is the no slip boundary condition that I mentioned on the first page of this thread. A typical laminar flow velocity curve looks like this.
In turbulent flow this orderliness starts to break down and plug flow starts to form. As the Reynold's number increases the flow velocity curve flattens out and becomes less parabolic. This means that fluid near the wall can have the same velocity as fluid in the ceter of the pipe. This is why turbulent flow is sometimes referred to as plug flow, because there is a plug of fluid in the center moving at a uniform speed and a ring around it which is moving slower (which creates drag). A low Reynold's Number turbulent flow velocity curve looks like this.
At very high Reynold's Numbers, this curve is essentially flat meaning that there is a very thin ring of stagnant/slow fluid around the pipe wall and the rest is rocketing through the pipe. This thin filim is called the laminar sublayer and is in fact laminar. The flow will break turbulent just a little ways away from the pipe wall.
Ok, that is the preamble, now to the point. In a typical skimmer, the Reynold's Number will be far into the laminar flow regime which means that the fluid velocity changes a lot from near the wall to the center. Also, the flow velocity at the center in unlikely to be all that fast anyway. In a garden hose with 100 gph going through it the Reynold's Number will be far into the turbulent flow regime which means that most of the fluid is moving at the same speed and this speed is pretty healthy.
Air bubbles (or any bubbles it doesn't matter) have a critical flow velocity called the incipient flow velocity. This is the velocity at which they become entrained in the flow around them. All this is is the velocity at which the drag forces on the bubble are greater than the bouyant forces on the bubble. In our counter current flow examples with conventional wide skimmers we are nowhere near the incipient flow velocity of air bubbles in water so they will rise as if there were no flow at all. If we keep the same flow rate but narrow the skimmer to 1/2" we will far exceed the incipient flow velocity and sweep the air bubbles out the bottom of the skimmer. If we approach but don't exceed the incipient flow velocity we get what we are after, long dwell times with slowly rising bubbles.
I know I didn't supply all the equations for calculating flow velocity at a given distance from a pipe wall or for calculating the incipient flow velocity but I'm not sure we really need them. In order to approach the incipient flow velocity of air bubbles we would need INSANE flow through typical skimmers (greater than 3" in width). This will cause turbulent flow which will strip proteins back off the bubble walls which is what we are trying to avoid.
Just doing my part to keep the Advanced Topics thread true to its name.
Fluid flow velocity is a factor of 4 parameters: fluid viscosity {μ}, fluid density {ρ}, fluid velocity {v}, pipe diameter {D}. This neglects skin friction losses which in skimmers is trivial. These four parameters are used to calculate a dimensionless ratio or number called the Reynold's Number. It is dimensionless because all the units cancel each other out leaving only a ratio. This ratio is the ratio between the viscous forces and kinetic forces in a fluid flow situation. When viscous forces dominate flow occurs in streamlines which don't intermingle (think good drivers on a very packed interstate where no one is changing lanes). When kinetic forces dominate viscous forces, the streamlines start to swirl and intermingle (think 1,000 car pileup happening at 100 mph or if you've ever driven in downtown Rome...).
The Reynolds number is given by Dvρ/μ. Where the numerator is the magnitude of the kinetic forces and the denominator is the magnitude of the viscous forces. Reynold's numbers below 2,000 make up the laminar flow regime. Between 2,000 and 2,500 is a transitional flow regime. Above 2,500 is the turbulent flow regime.
In the laminar flow regime, fluid velocity changes significantly between the wall and the center of flow. At the wall the flow velocity is zero. This is the no slip boundary condition that I mentioned on the first page of this thread. A typical laminar flow velocity curve looks like this.
In turbulent flow this orderliness starts to break down and plug flow starts to form. As the Reynold's number increases the flow velocity curve flattens out and becomes less parabolic. This means that fluid near the wall can have the same velocity as fluid in the ceter of the pipe. This is why turbulent flow is sometimes referred to as plug flow, because there is a plug of fluid in the center moving at a uniform speed and a ring around it which is moving slower (which creates drag). A low Reynold's Number turbulent flow velocity curve looks like this.
At very high Reynold's Numbers, this curve is essentially flat meaning that there is a very thin ring of stagnant/slow fluid around the pipe wall and the rest is rocketing through the pipe. This thin filim is called the laminar sublayer and is in fact laminar. The flow will break turbulent just a little ways away from the pipe wall.
Ok, that is the preamble, now to the point. In a typical skimmer, the Reynold's Number will be far into the laminar flow regime which means that the fluid velocity changes a lot from near the wall to the center. Also, the flow velocity at the center in unlikely to be all that fast anyway. In a garden hose with 100 gph going through it the Reynold's Number will be far into the turbulent flow regime which means that most of the fluid is moving at the same speed and this speed is pretty healthy.
Air bubbles (or any bubbles it doesn't matter) have a critical flow velocity called the incipient flow velocity. This is the velocity at which they become entrained in the flow around them. All this is is the velocity at which the drag forces on the bubble are greater than the bouyant forces on the bubble. In our counter current flow examples with conventional wide skimmers we are nowhere near the incipient flow velocity of air bubbles in water so they will rise as if there were no flow at all. If we keep the same flow rate but narrow the skimmer to 1/2" we will far exceed the incipient flow velocity and sweep the air bubbles out the bottom of the skimmer. If we approach but don't exceed the incipient flow velocity we get what we are after, long dwell times with slowly rising bubbles.
I know I didn't supply all the equations for calculating flow velocity at a given distance from a pipe wall or for calculating the incipient flow velocity but I'm not sure we really need them. In order to approach the incipient flow velocity of air bubbles we would need INSANE flow through typical skimmers (greater than 3" in width). This will cause turbulent flow which will strip proteins back off the bubble walls which is what we are trying to avoid.
Just doing my part to keep the Advanced Topics thread true to its name.
crazzyreefer
New member
thank you chemE also I'm trying to come up with a DO number, I know there is variables, Lighting, algae...etc.. but even still I would like to calculate the DO into the CFM intake of the skimmer.is there a way to translate ORP or Redox To DO?
I believe we all have basic knowledge here, but I thought the Reynolds Number was Re=Ua/y
I probably wrong, this isn't my field, I just design the equipment.
I believe we all have basic knowledge here, but I thought the Reynolds Number was Re=Ua/y
I probably wrong, this isn't my field, I just design the equipment.
ChemE
New member
You could assume that the sea water is saturated with oxygen since saturation is still only in ppm. I'm not aware of a conversion between OPR and ppm DO.
That equation could be correct so long as Ua = kinetic forces and has units of viscosity {cP} and y was the viscosity of the fluid.
That equation could be correct so long as Ua = kinetic forces and has units of viscosity {cP} and y was the viscosity of the fluid.
is there a way to translate ORP or Redox To DO?
No, there isn't. It must be measured independently.
No, there isn't. It must be measured independently.
tinygiants
New member
The ratio is bombardment rate not air volume. 120 sec = 2 minutes for air dwell. You need 20 minutes for the water dwell to be in the correct ratio (acording to Escobal) of water to air contact time.
tinygiants
New member
Tube size will affect the bubble rise at a rate of 1.17 inches/sec - water flow inches/sec (theory). The larger the tube the slower the velocity at a given GPM.
crazzyreefer
New member
dwell time and air ratio are separate issues. a bubble doesn't need 20 min, only 120 seconds, in that 120 seconds it has fully formed and and the organics and oils have attached, at this point it is unlikely to burst in the foam chamber, you couldn't have dwell time of 20 min if you wanted to! the reasoning behind the saturation point is the air is the filter media, so the less air ratio the less foam, the less filtering, above the saturation point of 13% the bubbles are no longer stable and join with others, causing belching in the foam tube and larger bubbles offer less surface space for the organics to adhere to. Escobal has decided that a ratio of 10% is perfect, but the actual number is really 13%.
Randy,ChemE, thank you.... I kinda knew that, but I was hoping that I was wrong. So it will remain an unknown, and CFM input would have to remain adjustable. or is so small of a number that is not important.
Randy,ChemE, thank you.... I kinda knew that, but I was hoping that I was wrong. So it will remain an unknown, and CFM input would have to remain adjustable. or is so small of a number that is not important.
Similar threads
