Fluid Mechanics and Dust Collection

[Never Enough Money]

Anybody know the losses associated with a "T" connection vesus a "Y", say at angle theta? I'm sure this is a standard problem in a fluid mechanics class.

I notice the Lee Valley catalog ONLY carries the "Y". Woodcraft carries "Y" and "T". Knowing the reputation of the Lee Valley staff, I suspect they do not carry the "T" because of the large vacuum loss around the sharp corner....and, therefore, refuse the sell it. Just a hunch.

I want to use "T" because they fit my layout better..but...hey it's nice to have strong suction forces too.

[Tyke]

Takes me back a long time to when I did these calculations at university.

I found a decent paper from the radon mitigation industry.

20drop in piping 201996.pdf

I did not expect this to include "Y" fittings, but the conclusion for the typical 4in DC duct is a "T" was about 22 ft of equivalent straight pipe. i.e. pressure drop is the same as having 21 addition feet of pipe. A sharp 90 deg (typical fitting) is almost 6 ft of equivalent straight pipe.

The radon installations are low flow compared to a DC.

In general, avoid "T" fittings. The "Y" is going to be somewhere between the 90 and the T, but closer to the "T".

Maybe someone else will post more accurate factors.

Dave Paine.

[Sierevello]

Fluid Mechanics deals with this in a number of ways. The first is through turbulent flow or laminar flow of the fluid through a pipe or valve. Laminar is "good" and turbulent is "bad". Laminar flow in fluid mechanics means that the fluid is flowing in smooth layers or laminae. The highest velocity in laminar flow is at the center of the fluid stream and the velocities of each respective layer is linear. At the walls of the pipe the velocity is 0 because of friction. Friction varies from pipe to pipe and fitting to fitting.

To determine the flow characteristic you must utilize the Reynolds Number, Darcy's Equation, Moody's Diagram, and the K Factor. Darcy's Equation is used to calculate head loss due to friction in pipes for both types of flow. Moody's Diagram uses the Reynolds Number and the relative roughness of the pipe to determine the friction factor. The "K" Factor is also known as the the Constant of Proportionality and is used to calculate the energy losses in valves and fitting such as tees, elbows, and bends.

So I do not think that this is as simple as you think and you are comparing apples and oranges. The viscosity of the fluid in these systems is also taken into account. With standard air there really is no "viscosity" although the density of the air changes with the amount of things that are in the air such as more oxygen than nitrogen etc.

If you utilize the Y fitting the flow of air will be much smoother and less turbulent than if you use a T fitting. With the T fitting the air molecules run right into a wall and then need to be reaccelerated. With the Y fitting, even though there are frictional losses, the deceleration of the air molecules will be much less because the change in direction is more gradual. If you use your car for example: Would you rather be traveling 40mph and come to a T in the road and try to turn left or a Y? With the Y the directional change the car would require less braking and thus less time to reaccelerate to your travel speed. With the T you will almost have to make a complete stop.

I hope this did not confuse you but fluid mechanics do not apply here from an engineering standpoint. However, the theories do. Let me know if you need any more help.

[Morris Dovey]

| I hope this did not confuse you but fluid mechanics do not apply | here from an engineering standpoint. However, the theories do. Let | me know if you need any more help.

What area of study (if not fluid mechanics) deals with air flow in a plenum?

Can you suggest some web sites that could provide some good background in optimizing airflows? My interest is in maximizing efficiency of solar heating panels; with a secondary interest in design of dust collection systems.

-- Morris Dovey DeSoto Solar DeSoto, Iowa USA

[Lew Hodgett]

Yes I could but that was a long time ago, not now.

Let me try a little analogy.

Remember a couple of things:

1) Crap doesn't run uphill and neither doe dust for very long. 2) There are no "T" fittings made or used in a sewer stack, only "Y" fittings, and with good reason.

Dust collectors and sewer stacks are a lot a like. Neither one handles sudden changes in direction well.

Enjoy you D/C equipped with "Y" fittings.

Lew

[Gerald Ross]

The ones I use are not a true "Y" but a straight with a 45 degree branch. Another 45 and you have a 90 total, but it is more gradual.

[Never Enough Money]

Thanks for the answer. I know this is not a simple problem but I suspect thee are rules-of-thumb that apply. I suspect theres a cosine of theta involved, too.

[Never Enough Money]

The link didn't work for me. I appreciate your answer, though. As I mentioned in another reply, I suspect there's a cosine of theta involved (or maybe a sine of theta)....So a T causes a 22 foot pressure drop, then a Y of angle theta would cause, a guess:

Pressure drop = 22*sin(theta). So at 90 degrees (a T), the dropis 22 feet. At 0 degrees (straight pipe), the loss is zero.

So at 45 degrees, the loss is 22*sin(45)=22*0.707 = 15.5.

I know this is back-of-the-envelop, but is it even close?

Tyke wrote:

[Never Enough Money]

Seems like the two 45's would sum to 90 in losses - so they would be equivalent. I guess, there may be some non-linearities with the 90, though that cuase linearity not to work at high angles (90 degrees for a T).

Gerald Ross wrote:

[Mike M]

LOL, your responses are good, shit doesn't flow anywhere but downhill

[Leon]

Do you think a right angle turn on to a freeway would be better than one merging at angle that is closer to parallel?

[Never Enough Money]

No I don't think that and what makes you think I do? However, there are conditions where a right angle turn makes senses.

My question is not whether "T" is better than a "Y" but by how much.

Leon wrote: [snip]

[Charles Spitzer]

they don't though. there's also inertia in fluids that is much less in aerodynamics, at least at the speeds of the air that are under discussion.

[Leon]

I think it would be easier to maybe buy each and see which works out best for you. To know the percentage of how much better you need to know at any given point, temperature, humidity, barometric pressure and volume of air moving through the intersection.

[Never Enough Money]

For relative comparisions when temp, humidity,etc are held constant but the angle of the "Y" is varied, there's got to be a rule of thumb -- say vacuum loss is proportional to the sine of the angle to the cube power....or something....

Le> > No I don't think that and what makes you think I do? However, there are

[Dave Hinz]

But how does the grounding wire change the flow? Clockwise, or counterclockwise twist to the strands?

[B a r r y]

Large chips get stuck on it, ending the flow altogether.

[Never Enough Money]

If your chips are that large, you need a new blade.

B a r r y wrote:

[Never Enough Money]

Engineering estimate: Let's assume the wire is coiled in such a way as to represent a cross-sectional area (when looking through the tube like a telescope) of 1 square inch. That seems hign but let's go with it.

Now a 4 inch diameter tube has a cross sectional area of 12.5 square inches. Thus, I claim approximately 8% of the are is lost so there's an

8% loss of vacuum every foot.

My 1 inch is probably high, it's probably closed to one tenth of an inch so the result is closer to 0.8 percent.

I know nothing about fluid mechanics except that my roommate used to spend entire nights doing the homework (I was in electrical engineering, he in mechanical). However, I'll bet that my super simple minded approach here is not that far off.

Dave, sorry for tak> > > For relative comparisions when temp, humidity,etc are held constant but

[Dave Hinz]

OK, so far so good.

No, that would be a loss of _flow_, not vacuum. Unless you're poking holes in the side of the pipe, you're not losing vacuum.

If that. Maybe .01" cross section. However - fluid dynamics are a fluid and dynamic field of study, and just as you'd expect, things behave in unexpected ways.

It's a restriction, but it's more than that - it will introduce turbulence which further impedes flow. Probably.

No worries there, I assure you.