hairpin turns vs. right angles

In terms of flow resistence in a forced circulation boiler, is there any difference between a hairpin turn as compared to two right degree turns with a length of pipe in between?

Richard:

Here's a simple mind-game to answer your question (please refer to the attached, crude graphic which will illustrate my remarks, I didn't aspire to graphical excellence here so no remarks about that, plz!):

First, let's assume you're trying to achieve the same goal with 2 different layouts, i.e., the "entry" and "exit" legs of the hairpin are spaced apart with the same dimensions.

In example "A" the bend to connect the 2 legs is continuous and in example "B" the legs are connected with smaller-radius bends and there is a straight leg between which is perpendicular to the entry and exit legs.

The obvious answer is that the larger radius of the continuous connector is much more beneficial to improved flow; not only is the radius larger in "A" but it involves only half the transitions from straight to curved.

I think the answer is simple, make your own observations.

Edit: Of course the magnitude of the curvatures is also important, for large radii the difference would become irrelevant. Question is, what is the transition into irrelevance?


Bill



Edited 1 time(s). Last edit at 01/13/2013 06:23PM by Bill Hinote.

Bill
It was an effort just to figure out how to word my question, but your quite addiquet sketch will help move it along. So... assume the radious of both tubes examples to be that of B. as if connecting up two 90 degree pipe elbows. Connect example A with the shortest nipple possible. Connect ex. B with a 2 ft. section of pipe. The question is, does the added 2 ft. of pipe in between the elbows do anything to reduce friction? I am assuming the streight section between elbows is of no value but wanted another opinion.



Edited 1 time(s). Last edit at 01/13/2013 07:02PM by richard orr.

richard orr Wrote:
-------------------------------------------------------
> Bill
> It was an effort just to figure out how to word my
> question, but your quite addiquet sketch will help
> move it along. So... assume the radious of both
> tubes examples to be that of B. as if connecting
> up two 90 degree pipe elbows. Connect example A
> with the shortest nipple possible. Connect ex. B
> with a 2 ft. section of pipe. The question is,
> does the added 2 ft. of pipe in between the elbows
> do anything to reduce friction? I am assuming the
> streight section between elbows is of no value but
> wanted another opinion.

Richard:

The simplest answer I can offer is, it's a real world! In the end we just find ourselves having to find a way to pipe the whole mess up and keep it from leaking excessively.

If you're lucky enough to have a really high-performance steam package then you can try to refine the delivery system. I doubt that one of a hundred of us in this forum has even found this refinement necessary.

FWIW, Bill

In any length of pipe or tube there is resistance to flow.

A hard right angle will have the greatest pressure drop. If you think about that case it is a dead head into the corner and the pressure in that volume is driving the flow out the exiting tube.

Where you have a smoothe transition into and out of the corner you have a continuous flow.

In any case: a change in direction takes energy. A change increasing altitude requires energy. A change decreasing altitude puts energy into the flow.

To be specific about your question, assuming no change in altitude and the radius are the same. Two 90 degree smooth transitions with a straight path between them as compared to a 180 degree smooth transition will have an additional flow resistance only if that straight path case has an overall greater length. I.E. A 6" straight section to a 90 degree bend to a 6" straight section to a 90 degree bend to a 6" straight section. should have the same flow resistance as 9" straight section to a 180 degree bend to a 9" straight section.

Andy

Andy
Thank you for the feedback. I am over-thinking as usual yet still curious weather there is some nuance of difference between the flow of the two configurations described. I will be using 1/2 in. inside dia. pipe in a once-through boiler. I understand that the back pressure of steam bubbles in a low presure sceanario will be the main facter to friction and probably the velocity of the feed water is slow enough so as to not be significant.

What actually matters is the specific volume of the total steam/water mixture jumps very quickly; this means the steam/water mix travels at much higher speeds than expected and this is where the resistance to flow comes from....

For example, at 200 psia for quality at 10%, the specific volume is 13.5x of that of water at that temp and pressure if I wrote the program correctly. So a piece of 1/2" schedule 40 pipe carrying 2 gpm would go from 2.1 ft/sec velocity to 28.5 ft/sec. I've not come up with a program to compute the resulting pressure loss, but my guess is that it is going to climb a lot with those high velocities.

Anyone with suitable algorithms for computing two phase steam/water pressure drop, chime in...

- Bart

----
Bart Smaalders [smaalders.net]

Lots of useful calculators on here Bart
[www.pipeflowcalculations.com]


Mike

Thank you, Bart. I'm on board with that. What I am pondering probably doesn't amount to a hill of beans in terms of boiler performance... just an exersize in curiosity, but it does seems to me that up to a certain velocity the interplay between a right angle elbow and a strait section is that of turbulent flow and then flow SETTLING DOWN to laminer until it hits another elbow. In the case of two elbows making a hairpin turn, it is turbulance - turbulance and then laminar. I am wondering if there is, in between those two elbows, something more then just the subtraction of the pressure drop represented by both elbows but rather a nuance of a compouinding negative as in " the sum of the total equaling LESS then it's parts" (Probably no body cares but just chuckle to yourselves and leave me to my fun)



Edited 6 time(s). Last edit at 01/15/2013 04:07PM by richard orr.

I don't claim to know anything by personal experience in this matter. Just commenting out of curiosity. I would think that turbulent flow is desirable. I would guess that the loss due to pressure drop in the tube is more than repaid by increased heat transfer.

Kerry

richard orr Wrote:
-------------------------------------------------------
> Thank you, Bart. I'm on board with that. What I am
> pondering probably doesn't amount to a hill of
> beans in terms of boiler performance... just an
> exersize in curiosity, but it does seems to me
> that up to a certain velocity the interplay
> between a right angle elbow and a strait section
> is that of turbulent flow and then flow SETTLING
> DOWN to laminer until it hits another elbow. In
> the case of two elbows making a hairpin turn, it
> is turbulance - turbulance and then laminar. I am
> wondering if there is, in between those two
> elbows, something more then just the subtraction
> of the pressure drop represented by both elbows
> but rather a nuance of a compouinding negative as
> in " the sum of the total equaling LESS then it's
> parts" (Probably no body cares but just chuckle
> to yourselves and leave me to my fun)

Kerry is right - the point is that you want to be well turbulent in the long stretches of pipe - what the "wets" (fluid & thermo engineers) called fully developed flow - in a pumped circulation boiler. To do this, keep your Reynold's Number above 8000 or so. In the situation I was working on, 2 GPM in 1/2" pipe means 10K or so - well turbulent.

There really isn't a good model for two phase flow through fittings; there are a lot of approximations you can find in the literature. I think you could learn something with a manometer, some pipe + fittings and a flow gage (or stop watch and bucket!).

- Bart

----
Bart Smaalders [smaalders.net]

Bart

In reading about rennolds numbers, and speaking about them in the most general of terms, it seems that the basic principal is making sure that the greatest amount of molacules of water impinge upon the inside of the pipe as is reasonaly possible. I say "reasonaly" because I assume that on the other side of the equasion there are losses due to friction as velocity increases. I understand that in the case of a lamont style boiler the velocity is an insurance policy against tube burn-out in a fluctuating boiler. But that aside and assuming a boiler in a steady state, it would seem advantgous to keep water velocity within a certain range, a "sweet spot" to keep gains and losses in balance.
I should clearify that I understand recirculating water in a lamont opperating at high pressures requires minimal energy from the over-all system.
My thoughts all generate around designing a mono-tube to opperate at around 150 p.s.i.g. and therefore having much more back pressure.



Edited 4 time(s). Last edit at 01/17/2013 02:26PM by richard orr.

For a monotube, the ciculation is much lower - e.g instead of 5x the steam output rate, it is 1x... but you still want to keep pressure drops reasonable - so you start w/ small diameter tubing and then increase the size to cut the speed and pressure drop as the specific volume increases. On the other hand, since you are using a positive displacement pump, getting even several hundred psi is not difficult. In our previous example, we'd be doing a bit less than 1/2 gpm. Even with a 500 psi. pressure delta in the boiler, that's less that .2 hp to drive the pump.

- Bart

----
Bart Smaalders [smaalders.net]

Bart
I am confused here when you say "previous example" But to extrapolate from the basics of what I am infering from your posts: Calculate back from the steam demand as set by the engine particulars and operating pressure. Size the pipe to achieve a renolds number arounnd 8k, and increase diameter (at some point along the generating tubes) to minimise pressure drop while maintaining renolds number at a constant. And the relationship of a displacement feed pump is not much effected by the amount of pressure drop taking place in such a scenario.
I am realizing that what I need to do is get a book specializing in mono-tubes so that I can answer some of the basic theory questions myself. I've read opinions on this forum that there should be a special 101 section for the unproven neophyts to post our questions. - thats got my vote ... a little play-pen-of-the-basics that we can bump around in without getting lost and the more experienced can drop by from time to time and administer spoonfulls of informational pablum. Could call it " Questions from the playpen "
. However, that being said, here are a few more questions about boiler basics: (1) Do the renolds numbers apply as criticly to the economiser section? (2) Is there formula for calculating from enthalopy tables to (ball park) where the diameter of the generating tubes should be increased to minimise pressure drop?
I've read in past posts responding to questions such as these, and the answer is usualy that it depends on certain variables. It would be helpful, though, to offer a list of of those variables and some theoreticle overview of their relationships to begin zearoing in on. --- Gotta get that book ---!



Edited 4 time(s). Last edit at 01/18/2013 11:55AM by richard orr.

I've not done the calculations to determine the boiler heat balance; I'd love to see a worked example myself. Basically, your gases start out very hot (you can calculate gas temps assuming 0% excess air and knowing fuel composition); as they go by each tube they lose heat and the tube gains heat. In the case of a Lamont boiler the tube is approximately constant temperature (since both water and steam are present), in a monotube the input water is whatever comes out of the economizer, and it climbs in temp. until it starts to form steam, is relatively constant temp. until the steam reaches 100% quality and then the temps start to go up again. Nice calculation for a computer....

I would work on keeping flow turbulent in the economizer - it just requires using smaller diameter tubing, and the pressure drop is minor since there's no steam formation.

If you're building a marine monotube, I'd build a "flash overfeed" version - bring the coil out of the boiler casing about 3/4 of the way through, and connect a steam trap there. Them, have an adjustable displacement pump, and adjust the pump displacement to provide a small flow out of the trap back into the hotwell (or condenser). This insures that the steam temps stay reasonable and keeps the engine from getting a slug of water.


- Bart

----
Bart Smaalders [smaalders.net]

Bart
I was thinking about doing something similar to what you described as flash over feed. In your version, another pipe goes from the trap and back into the boiler as super heat? I was also thinking about recirculating lamont style only re-circulating through a smaller portion toward the end of the generater tube and rather then recirculating at 5x, just recirculate at 2x.



Edited 1 time(s). Last edit at 01/19/2013 09:55PM by richard orr.

I think there was mentation of Doble formula in another thread. But for a modern steam vehicle we should start by looking at requirements. I am not so sure we can predict the evaporation zone. The one thing that I have spent a lot of time looking at is the range of steaming rate required.. There is a necessary vary large range of steaming rate requirement. The aerodynamic square law of speed to force, S = const F², turns into a cubic relation to power requirement. The cubic relation starts some ware from 15 to 20 MPH. Below that there is close to a linear speed to force relation or a square relation for speed to power. So just looking at speed from 20 tp 80 MPH we have a 4 to 1 speed range. That translates to a 64 to one power range. Power is a direct relation to steaming rate. So we are looking at greater then a 64 to 1 steaming rate requirement.

As an example say we require x HP to maintain 20 MPH. At 60 MPH we have tripled the speed requiring 9 times the force to over wind resistance. and at the same time we are traveling 3 times the distance in the same amount of time. So we must be do 36 times the work per time as at 20 MPH. Power requirement increased with the cube of the speed. x HP at 3 times the speed 3*3*3*x HP.

Then figure the valves opening and closing generating intermittent demands in flow and producing pressure waves propagating through the system.

When I last talked to J. Carter. He said that he was thinking of trying to control the evaporation zone with movable baffles. He wonted to get a better control of the supper heat temperature going to the engine.

Here are two observations. The first one comes from Jim Tangeman and it is just as important to have pull or resistance on the pipe being bent in a hair pin bender as it is in a coil winder. The pulling through a clamp starts to stretch the metal in the tubing and thus when it is bent it does not flatten out. This is one of those non-obvious things. The second observation comes from Jim Crank and the square boiler that he inherited from Lear that was in his Land Speed record car, the one that is now in the museum in Reno that used to be Harrah's museum. Jim said it was a real steamer and he opined that the sharp turns in the corners caused the water flow to swirl and thus scrub steam bubbles off the tubing walls. Nowthen, few of us have been inside a tube watching what happens but all of us have wondered what was going on inside the tubing of a monotube steam generator. Tom Kimmel

Tom
I would bet that Jim is right about the swirling of water around hairpin turns.I can readily see it in my mind's eye as it seems to be the nature of water. In the case of a screwd pipe elbow, I take it to be documented that there is turbulance rather than swirl, but still with similar scrubbing.
The late Cidny Climent (Steamboats and Modern Steam Launches) used screwed together pipe for his mono-tubes in his shop. He claimed that with proper sizing, control was no problem. He also claimed that a low pressure mono-tube was no problem either. He was speaking in terms of use in boats, but did say that mono-tube control in a vehical was an entirely different animal.
In the case of an experimental mono-tube, sections of pipe could be easily added or subtracted. (Making a case for screwed togather boiler here) (for my boat)
In fact, I think at this point I should redirect the focus and explain the point of this thread or where it is going in the first place, what it is guided to solve towards: I am intending to build a mono-tube boiler for my sidewheel steamboat and another for my shop. All my focus is directed to the design of a screwed-together mono-tube boiler - lightness is always a big facter in a boat. I think a succesful design of such would open the possibilitys of steam use to a wider group of would-be steamers because it would be the least expensive and safest type of boiler to make for those of limited skills.
I am hooked on steam as I believe the rest of you are - hooked as in it is a passion and a percieved solution to looming energy problems we will all soon have to solve for. Transition will either be gradual or it will be sudden. It seems smartest to solve for worst case sceenario.

This thread began with a particular questioni in mind, It was about the relationship of one right degree elbow to another in order to obtain a 180 degree turn. Seems absurdly simple enough in itself and hardly reason to give pause .Perhaps and probably it is just as Bill H.said - To paraphrase, "Just screw the *&^%* thing to gather and get a fire under it". However, I was thinking in terms of something I read on another thread. It was about the resonancy created in expansion chambers in two cycle engines. The author was relating that technology towards steam engines and how the sizing of the valve chamber could be used to create a sound wave to help speed up the tiansfer of steam past the inlet valve. It got me to thinking about weather there was a resonancy happening at a given velocity of steam within a right degree elbow and weather the length of the tube that the steam exited to was a facter to be concidered. Perhaps there is nothing going on within that elbow besides plain ol' turbulance, too chaotic to be concidered or worked with and it is simply a matter of just getting on with the job of getting the steam/water mix past the elbow. Then again, there is a lot of bloody bleeding, blooming friction/resistenc in each elbow, enough that ( If one were commited to the concept of a screwed-togather boiler as I am) it would justify a little thought in this matter. Am regretful that at this point that I have nothing to offer up other then the question itself., but on one level it is all in the name of fun anyway or as Tom said, we would all like to see what's going on inside those tubes. Blorkaroondus flamoongus ya'll.



Edited 7 time(s). Last edit at 01/21/2013 03:32PM by richard orr.

All files from this thread

File Name	File Size		Posted by	Date
hairpin comparison.jpg	21.6 KB	open | download	Bill Hinote	01/13/2013	Read message