Not sure how well this will post. It looks like the subscript and superscript will be lost. I still need to foot note it and such. We'll see, I hate doing footnotes. If its good enough, then I will.
Give it a read if you can manage. Any comments or criticsm would be appriciated.
Additionally, no the paper itself isn't smoking a pipe. There's a character limit on the title, and writting it gramatically correct wouldn't fit.
Thanks,
Nick
Smoking quality is predominantly a function two things: cool smoke and dry smoke. These two facets can be broken down into what creates the desired results. Cool smoke is a function of effective heat transfer, which deals with the conductive and diffusive properties of briar and the stem material.
A dry smoke is a function of several factors. Leaving tobacco moisture content aside, the foremost factor is the engineering of the pipe. A well engineered pipe will create a laminar flow of the smoke stream. That is to say the stream of the smoke will consist of parallel concentric channels of smoke particles within the draught hole. Laminarity and turbulence are measured by the Reynolds (Re) number. The Reynolds number is determined using the following formula:
Re = D*v*r /m
or
Re = D*G/m
Where
D = Pipe length
v = velocity
r = density
m = absolute viscosity
G = mass velocity
Fluid streams with a Re value of 0 – 2300 are said to be Laminar. A laminar flow is a condition where the fluid flows in parallel streams or streamlines at differing velocities. These streamlines never cross. In a circular pipe, the centermost streamline is flowing at the highest velocity, while the streamline closest to the pipe wall has a velocity near zero. Turbulent flows are characterized with an Re number of 4000 or higher. A turbulent flow is a chaotic flow, with little or no pattern to the streamlines. Turbulent flows often are made up of swirling vortices. Flows with Re numbers greater than 2300 and less than 4000 are said to be Transitional, having characteristics of both laminar and turbulent flows.
Moisture in the smoke stream is a naturally occurring phenomenon. And in a laminar flow, this moisture does not present a problem. However, if the flow is turbulent, this moisture will coalesce into larger droplets and collect in the pipe.
There are potentially five points within a pipe where turbulence can be created. First is the where the draught hole meets the tobacco chamber. Flows can become turbulent when subjected to a sudden change in direction or a constriction in pipe diameter. Additionally, roughness of the draught passage walls, flaws at the mortise/tenon joint, the bending of the stem and at the rapid change in pipe size at the stem button can also cause turbulence.
While to date I have not found a mathematical formula to isolate turbulence or eddies created by such junctions, one can calculate the pressure loss or head loss generated. This is done through what is called the Minor Loss Coefficient. The formula for calculating this coefficient is listed below:
ploss = j 1/2 ρ v2
or
hloss = j v2/ 2 g
Where
j = minor loss coefficient
ploss = pressure loss
hloss = head loss
ρ = density
v = velocity
g = acceleration of gravity
In our present examination, we will account for a number of loss coefficients at the same time. To come to a single loss coefficient, these loss coefficients are multiplied and the resultant number is then used as a whole to figure the overall pressure loss. Using the pressure loss generated by the change in the pipe, one can find the velocity. The new velocity can then be plugged into the Reynolds Number formula and the Re number derived.
An important question in this exercise was whether the bowl/draught passage junction is a bend or an inlet. The junction has characteristics of both. It is an inlet in that the draught passage meets a larger chamber, but it could also be considered a bend if the smoking pipe is considered one whole pipe. Details of how these two attributes would be addressed are discussed below:
An inlet is generally defined as a passage leading to a cavity. In this case the cavity which the inlet leads to is the smoker’s mouth. Shown below are illustrations of several types of inlets.
Minor loss figures for such inlets are usually arrived at through experimentation. However, generally accepted coefficients for standard inlets have been established. These figures are seen below:
Inward Projecting: 0.78
Square Entrance: 0.5
Rounded Entrance:
R/D = 0.05 0.25
R/D = 0.2 0.1
R/D > 0.2 0.05
Where:
R = the radius of the bend from the center point of an imaginary circle to the centerline of the pipe
D = the diameter of the pipe.
It should be noted that on a pipe with a square entrance R/D = 0.
Generally, an inlet is a straight passage leading to the chamber it is connecting. However, in our smoking pipes this is not always true. These passages can have bends at any number of angles. To that end, the coefficient needs to be modified to take into account the varying angles with which the draught passage intersects the pipe bowl.
A pipe bend is relatively self explanatory. A bend is simply where a pipe changes direction. As before, loss coefficients for pipe bends are usually arrived at through experimentation. However, generally accepted coefficients for standard bends have been established. These figures are seen below:
90o bend, sharp 1.3
90o bend, with vanes 0.7
90o bend, rounded radius/diameter <1 0.5
90o bend, rounded radius/diameter >1 0.25
45o bend, sharp 0.5
45o bend, rounded radius/diameter <1 0.2
45o bend, rounded radius/diameter >1 0.05
Additionally, we are able to arrive at the loss coefficients of varied bends mathematically using the following formula:
ja = j 90√a/90
In the end, deciding whether the bowl/draught passage junction was an inlet or a bend with a contraction was a practical decision. More data available on calculating the effects of a bend and contraction than there is on an inlet with an intersection angle different than 90 degrees. So, for the purposes of this paper, the bowl/draught passage junction will be considered a bend with an abrupt contraction
As noted above, a constriction of the pipe’s diameter, from bowl to draught passage, can create eddies or turbulence. The pressure losses from this transition can be measured using the minor loss coefficient. The MLC for a pipe constriction is determined using the following formula:
j =0.5 - (v1 / v2)2
In this case the velocity of both sections of the pipe needs to be known in order to determine the pressure loss. The initial velocity can be determined using Velocity of a fluid from one section of a pipe to another section can be determined using the mass continuity equation, which states:
r1a1v1 = r2a2v2
Where:
r = fluid density
a = area
v = velocity
The second point at which turbulence can be generated is the draught passage itself; specifically, the walls of the draught hole. Rough walls will create friction against the smoke stream, which can in turn generate turbulence. Wood pipes are generally accepted to have a roughness of between 600 to 3000 * 10 -6 feet or 0.2 to 0.9 mm For the purposes of this investigation we’ll pick a mid number of 0.4 mm. To calculate the relative roughness of a pipe, one uses the following equations.
For laminar flows, the roughness can be essentially discarded, and the friction factor is solely dependant on the Reynolds number. T For laminar flows the friction factor can calculated using the formula listed below:
l = 64/Re
For a turbulent flow the calculations are substantially more complicated.
l =0.25/(log10(e/3.7D+5.74/Re 0.9))2
Where:
l = Friction factor
e = e/D
e = absolute roughness
D = pipe diameter
Re = Reynolds Number
The third point at which turbulence can develop is at the mortise tenon joint. At this point two things can create turbulence. First, a gap between the mortise wall and tenon face will create an open space or plenum, where turbulence will develop. This situation is akin to having two significant changes in pipe diameter in rapid succession. The second potential point of turbulence or eddies develop is if the pipe changes diameter from the pipe shank to stem. Again, this issue was discussed above.
Fourth, a bend in the stem of a pipe can create eddies or turbulence. This bend occurs in virtually all non strait pipes. However, this bend is usually gradual, and it’s likely that very little agitation of the flow occurs here. The effects of this can be seen by applying a minor loss coefficient as noted above.
Lastly, turbulence can be created at the button on the stem. If the draught hole radically changes shape, affecting the flow of the smoke stream, turbulence can develop and droplets of moisture can accumulate. However, it is also likely that the effects of this are negligible, and hence will not be considered here.
The experimentation explored in this paper is confined to mathematical measurements. At the point of writing, I lack the facilities to run physical experiments.
The baseline data for this paper was gathered from various archived documents from the larger tobacco companies, predominantly Phillip Morris and Lorillard. These data are listed below:
Puff volume: 50 cc
puff duration: 1.2 sec
rate: 11 puffs/min
flow rate: 2430 cc/min = .6420 GPM = 145.7982 L/Hr
particle size in pipe smoke: 0.6 – 0.7 micron
static burn rate: 47 mg/min
fire holding capacity: 175 seconds
puff resistance before lighting:2 mm H2O (Resp<1) = .1471 mm HG = .0002 atm
puff resistance during smoking: 10 – 20 mm H2O = 1.103 mm Hg = .0026 atm (15 used for conversions)
Finding data on the physical properties of tobacco smoke proved harder than one would anticipate. Because of this, air, which makes up roughly 70% of tobacco smoke, was used as the fluid in these calculations. Air has the following properties:
Fluid viscosity(r) = 1.88 * 10-5
Absolute viscosity(m) = 1.31 mg cm3
Intro to a pipe smoking paper I'm writing. Opinions wanted
Heheheheheee
Well, thank you both for reading it. This is the most complex and taxing part of the paper. I presume for your responses that I could do some more clarifying? Make things a bit easier to understand? I've been messing with it for so long that it just makes sense to me. Does it flow OK?
If its good enough, I plan to send it into P & T, PSE and maybe some scholarly journal. We'll see.
I plan to delve into thermodynamics after I get this one polished off. Too much fun.
Well, thank you both for reading it. This is the most complex and taxing part of the paper. I presume for your responses that I could do some more clarifying? Make things a bit easier to understand? I've been messing with it for so long that it just makes sense to me. Does it flow OK?
If its good enough, I plan to send it into P & T, PSE and maybe some scholarly journal. We'll see.
I plan to delve into thermodynamics after I get this one polished off. Too much fun.
BTW Pooka, I think using autocad, or at least some 3d rendering program to design pipes is a cool idea. I've already designed up one or two that way.
But, I'm a dork and I know it. I read fluid dynamics for fun. I think about vortices, energy loss and condensation as I blow the leaves off my yard and wow my friends at parties with my obscure factoids about the viscosity and density of air.
LOL
And you said you were the Wyle E Coyote of the pipe making world!
One of the cool things a prof at OSU suggested doing was making a life size or scaled up pipe out of glass to actually see where turbulence developed. Too cool!
But, I'm a dork and I know it. I read fluid dynamics for fun. I think about vortices, energy loss and condensation as I blow the leaves off my yard and wow my friends at parties with my obscure factoids about the viscosity and density of air.
LOL
And you said you were the Wyle E Coyote of the pipe making world!
One of the cool things a prof at OSU suggested doing was making a life size or scaled up pipe out of glass to actually see where turbulence developed. Too cool!
Thank Pooka! I'm sure I'll be bugging you when I get around to thermodynamics. 
Scott,
Well the goal is just that I enjoy doing it. The idea that got me started is that pipe design inherently created the turbulence that is so bad for smoking. Turbulence generating coalesced moisture in a pipe, and hindering heat disapation. My goal for this specific section is to be able to measure the amount of turbulence generated in varying pipe designs.
What I'd like to be able to say is that a pipe designed "this way" will be a better smoker as far as turbulence generated.
Thermodynamics is another section all together which I haven't got into yet.
Scott,
Well the goal is just that I enjoy doing it. The idea that got me started is that pipe design inherently created the turbulence that is so bad for smoking. Turbulence generating coalesced moisture in a pipe, and hindering heat disapation. My goal for this specific section is to be able to measure the amount of turbulence generated in varying pipe designs.
What I'd like to be able to say is that a pipe designed "this way" will be a better smoker as far as turbulence generated.
Thermodynamics is another section all together which I haven't got into yet.
Well, at the risk of boring the heck out of the readers of the forum, I'll add some thoughts. Heck, if you're still reading this thread, you can only blame yourself.
It seems to me that you have taken the approach of decoupling the two problems (momentum and heat transfer), probably with the approach of attributing moisture ("wet smoke") to momentum transfer (or fluid mechanics), and heat ("hot smoke") to heat transfer or thermodynamics. Is that the case?
I'm not so sure you can do that. For smoke flowing within the pipe shank, heat loss through the shank wall would benefit you in one sense, allowing for the smoke to be cooler when it reaches the button. But it will also cause water vapor to condense.
Why did you spend so much effort on trying to quantify the pressure drops across the different junctions? From the data gathered you can already calculate the velocity of the stream in the straight section of pipe (in fact it looks like you already listed it above?). It looks like you are using properties of air (density, viscosity). Will this be good for something that consists of 30% (your number above) "other"? What about the temperature dependence of these properties? Won't you need to know the temperature profile in the shank to determine what kinds of changes you will see (i.e. if you treat air as an ideal gas, density will change as 1/T)?
I've never attempted to calculate the Reynolds number for this problem. Have you gotten an initial estimate (using a "standard" shank diameter)? What is it? If flow is laminar, wouldn't you then be interested in looking at the entrance lengths involved to determine if any turbulence involved in flow through an irregular junction is long-lived or not? The entrance length would depend on the pipe inner diameter and the Reynolds number.
Add to that the fact that flow in the shank will be pulsatile and not steady...
Seems like you got yourself a real tiger by the tail with this one.
-Scott
Did you know that the thermal conductivity of wood is different depending on whether heat is flowing with or across the grain?
It seems to me that you have taken the approach of decoupling the two problems (momentum and heat transfer), probably with the approach of attributing moisture ("wet smoke") to momentum transfer (or fluid mechanics), and heat ("hot smoke") to heat transfer or thermodynamics. Is that the case?
I'm not so sure you can do that. For smoke flowing within the pipe shank, heat loss through the shank wall would benefit you in one sense, allowing for the smoke to be cooler when it reaches the button. But it will also cause water vapor to condense.
Why did you spend so much effort on trying to quantify the pressure drops across the different junctions? From the data gathered you can already calculate the velocity of the stream in the straight section of pipe (in fact it looks like you already listed it above?). It looks like you are using properties of air (density, viscosity). Will this be good for something that consists of 30% (your number above) "other"? What about the temperature dependence of these properties? Won't you need to know the temperature profile in the shank to determine what kinds of changes you will see (i.e. if you treat air as an ideal gas, density will change as 1/T)?
I've never attempted to calculate the Reynolds number for this problem. Have you gotten an initial estimate (using a "standard" shank diameter)? What is it? If flow is laminar, wouldn't you then be interested in looking at the entrance lengths involved to determine if any turbulence involved in flow through an irregular junction is long-lived or not? The entrance length would depend on the pipe inner diameter and the Reynolds number.
Add to that the fact that flow in the shank will be pulsatile and not steady...
Seems like you got yourself a real tiger by the tail with this one.
-Scott
Did you know that the thermal conductivity of wood is different depending on whether heat is flowing with or across the grain?
Too cool guys.
Scott:
Pooka:
[OR, make your benchmark pipe and measure all other pipes against it, developing "briarodynamics" withyour own set of measurements and terminology. But that's no longer the paper you're looking at. [/quote]
Briarodynamics!! Love it!
So....you two want to colaborate with me?
Scott:
Ummm....well it was.It seems to me that you have taken the approach of decoupling the two problems (momentum and heat transfer), probably with the approach of attributing moisture ("wet smoke") to momentum transfer (or fluid mechanics), and heat ("hot smoke") to heat transfer or thermodynamics. Is that the case?
Really good point. I'm not sure how to address this. Perhaps in a another section or paper I could discuss how heat affects moisture development in a pipe. Really good point.I'm not so sure you can do that. For smoke flowing within the pipe shank, heat loss through the shank wall would benefit you in one sense, allowing for the smoke to be cooler when it reaches the button. But it will also cause water vapor to condense.
Partially just because I wanted to know and enjoy doing it. Partially because my first calculations showed a laminar flow. This just seemed wrong in my gut. So I emailed a few profs, who concurred with me, stating that though the final bit of the flow might be laminar, there are probably eddies or turbulence present, and the flow simply has time to relaminarize. By isolating each facet of the pipe, I think I can identify the eddies or turbulence. We'll see.Why did you spend so much effort on trying to quantify the pressure drops across the different junctions? From the data gathered you can already calculate the velocity of the stream in the straight section of pipe (in fact it looks like you already listed it above?).
Well...its as good as I could find. I tell you I dug through tons of documents, digging for properties of tobacco smoke or even smoke in general. Damn if I couldn't find a fricking thing. There were results of expiriments that had to have used the density and viscosity, but it wasn't listed. Drives me nuts. If you know of anything, I'd love to see it.It looks like you are using properties of air (density, viscosity). Will this be good for something that consists of 30% (your number above) "other"?
*grunt* It will? Damn! I didn't know that. I guess you're right, I will need to know these aspects.What about the temperature dependence of these properties? Won't you need to know the temperature profile in the shank to determine what kinds of changes you will see (i.e. if you treat air as an ideal gas, density will change as 1/T)?
A rough estimate of the Re is around 13600. Although this needs to be refined a good bit.I've never attempted to calculate the Reynolds number for this problem. Have you gotten an initial estimate (using a "standard" shank diameter)? What is it?
Absolutly!!! But it looks like th flow is turbulent. Still need to crank through the numbers though.If flow is laminar, wouldn't you then be interested in looking at the entrance lengths involved to determine if any turbulence involved in flow through an irregular junction is long-lived or not? The entrance length would depend on the pipe inner diameter and the Reynolds number.
Yea, thats nuts! I have not idea how to accomidate that. I've essentially approached the problem from a steady state flow, with the thought of getting the base line down first, and the adapting it to the pulsating nature if I can.Add to that the fact that flow in the shank will be pulsatile and not steady...
Pooka:
Yea, the glass of course would be less absorbant. I wonder how to account for this? Any ideas?Aaaaaand, if you're making a glass pipe, you're using a less absorbing medium.
I don't know. Would it make a difference??Aaaaaand, are you going to test consistently with a cold pipe or a warm pipe?
*grunt grunt* Yup yup.Aaaaaaaand, don't forget that the smoke is water vapor in the first place.
Drainage whole would definatly be a good idea, but would it effect the flow? Kind of making the results inapplicable? Of course that begs the question: would having more mosture present, because of a less absorbant pipe render the results bad? I dunno.You'll need to put a drainage hole in the glass, that's for sure.
Yep yep, just what I thought.I'd start with a straight pipe, like a plain old poker, make the drills perfect, with the most perfect fit you can, and use that as the benchmark. Then start varying your constructions for your test.
[OR, make your benchmark pipe and measure all other pipes against it, developing "briarodynamics" withyour own set of measurements and terminology. But that's no longer the paper you're looking at. [/quote]
Briarodynamics!! Love it!
So....you two want to colaborate with me?
LOL @ Tyler. Hey man, if you remember, this kind of topic came up on another board as well. If I recall it met with a large amount of indifference from just about everyone. Like I said, I think the topic bores the heck out of most folks. But you have a cheme background too, right? Admit it - you find it fascinating...