Thread: Air speeds in and around duct openings and blades (NB Invisible dust discussed)

Originally Posted by doug3030

BobL,

there appears to be an inconsistency here. Hopefully you casn explain it.

Look back at youur post #21 in this thread, At 50mm from the opening the airflow from the 4" is 9 and the 6" is 11.

In the latest post I saw the 4" with the bellmouth at 13 and thought initially that this is therefore better than the 6" in post #21, but then I looked again and the standard 4" has suddenly jumped to 11.5, when it was 9 in post #21.

The baseline reading (unmodified 4" pipe 50mm from windspeed measurement is not consistent over the two experiments which makes me ask "what was different?"

However the latest set of readings do give a great deal of hope to the 4" pipe gang, since if the 20% gain in windspeed is correct then the bellmouth will assist greatly to collect at least some more of the invisible dust.

Doug

Well spotted but I would treat them as two separate experiments.
The air speed for the 4" close to the opening should be greater than for the 6" as the cross sectional area is smaller so the air speed should go up.

There are two differences between yesterday and todays 4" ducting.
In the latest measurements the ducting is 104 mm in diam with one end having a smooth rounded leading edge.
Yesterdays test was for a square shouldered 112 mm diam pipe (not smoothed).

Also bear in mind that while it is important to capture dust at source it's not just just air speed along the midline line that matters but air speed integrated across a working volume. The 6" pipe might draw a lower air speed 50 mm from it's opening but it draws it across a much wider working volume. A $99 vacuum cleaner with a 50 mm ducting will likely have an even higher air speed at 50 mm but it has a much smaller capture volume.

Nonetheless Mic's bell mouth is well worth the effort. I think the flange I have is just the beginning and a more refined bell mouth design is likely to improve things a little more.

Originally Posted by doug3030

What I am getting at is that the shape of the end of the collector is determining where it picks up the dust. I am starting to see that the right shaped collector pipes could have a far greater impact on dust collection than most of us would have thought.
:

Correct, I don't think I need to even do the experiment.

It's not just collector openings but all of the ducting.

Originally Posted by BobL

Well spotted but I would treat them as two separate experiments.
The air speed for the 4" close to the opening should be greater than for the 6" as the cross sectional area is smaller so the air speed should go up.

I see where you are coming from! Maybe the graph at post #21 needs to acknowledge that as even at a slower speed the higher volume taken by the 6" pipe will be far greater that the 4" pipe, how would the existing graph compare to one with the same data but using the speed and area of the pipe to convert the figures to cubic feet per minute? I understand that this is not really an accurate measure of what is happening as the cubic feet per minute going up the tube is a constant, but maybe it would be a better way of more accurately depicting the difference between the volume up a 4" tube as opposed to a 6" one. Unless you know of a better way of course, Bob

Originally Posted by BobL

Nonetheless Mic's bell mouth is well worth the effort. I think the flange I have is just the beginning and a more refined bell mouth design is likely to improve things a little more.

Bob, I know you would not have seen my post above when you made this post but it looks like we are thinking the same way on this part. For example, how wide does a flange over the end of the pipe have to be before it effectively prevents collection from behind the opening?

Doug

Originally Posted by doug3030

I see where you are coming from! Maybe the graph at post #21 needs to acknowledge that as even at a slower speed the higher volume taken by the 6" pipe will be far greater that the 4" pipe, how would the existing graph compare to one with the same data but using the speed and area of the pipe to convert the figures to cubic feet per minute? I understand that this is not really an accurate measure of what is happening as the cubic feet per minute going up the tube is a constant, but maybe it would be a better way of more accurately depicting the difference between an airspeed up a 4" tube as opposed to a 6" one. Unless you know of a better way of course, Bob

It is very difficult to calculate the CFM from this data. The most accurate way is to actually measure it.

You will notice I am not measuring the air speed at 0 mm. This is because neither of the direct air speed sensors I have go above 30 m/s and the air speed is greater than this at this point. The pitot tube I have will measure up to 60 m/s but does not work properly unless the air stream is stable and it needs around 1m of straight ducting to stabilise the flow.

I do this by sticking 1m long pieces of test ducting in between the port and the rest of the ducting and then inserting my pitot tube into the middle of the test pipe to measure the air speed inside the pipe and then mathematically summing the air speed time incremental volume across the diam of the duct. It's a PITA measurement to perform.

Bob, I know you would not have seen my post above when you made this post but it looks like we are thinking the same way on this part. For example, how wide does a flange over the end of the pipe have to be before it effectively prevents collection from behind the opening?

Doug

this is a good question.

Originally Posted by BobL

Correct, I don't think I need to even do the experiment.

It's not just collector openings but all of the ducting.

I am not sure you fully got my meaning there Bob.

The combined attributes of the ducting do of course determine the volume of air that will pass through the system.

The shape of the end of the pick-up pipe determines exactly where that volume of air is collected from. It is our best interests to ensure that it is collecting as much of the dust laden air as possible at the exclusion of relatively clean air that may also be close to the intake.

I see this as two separate issues; firstly minimising drag in the ducting to maximise the volume of air passing through, and secondly, ensuring that the intake is shaped to ensure that priority is given to collecting the air with the most dust in it close to the source.

Doug

Originally Posted by BobL

Well spotted but I would treat them as two separate experiments.
The air speed for the 4" close to the opening should be greater than for the 6" as the cross sectional area is smaller so the air speed should go up.

OK, another thought comes to mind here about the graph at post #21, the speed did go up for the 4" intake on the later post, but why was the speed for the 4" lower than the speed for the 6" on that same graph?

Bob I know you took the 4" measurements later after I suggested it but did anything else change? I was thinking it may be due to turbulence as a result of a short length of 4" inserted into an otherwise 6" system but i didnt think that would make that much difference. Thoughts?

Doug

Originally Posted by HeadScratcher

Interesting concept but I'm with you Bob, the practical applications are likely to be limited. While it might work in theory I tend to think it could do more harm than good if an object deflects it. But worth playing with to satisfy curiosity.

Dear Headscratcher,
I think you are under the misapprehension that the push/pull "concept" is somehow theoretical, and that you missed the reference within my post. That reference was to a manual written by industry experts who have assembled, as part of it, a survey of the methods used in industry, the push/pull being one (and they do not give qualification to its applications and limitations) But you do sound very certain in your reply so if there is something you learnt in your several months experience with dust extraction it is important to share it.

Originally Posted by BobL

Walking through Bunnings this evening and I ran across some 4" floor ducting flanges in the DVW fittings section. The fittings are not unlike the bell mouth inlets that Mic refers to in his post earlier in this thread.

This is what they are supposed to look like in X-section


So I set up a test to compare air flow with and without the flange.
Here you can see what it looks like


The 100 mm ducting fits inside the flange so it leaves a ridge so I stuck the 100 mm ducting into a lathe and turned an (not very symmetrical) internal taper on the leading edge.

I measured the air speed for
1) the smooth blunt end of the 100 ducting
2) the internally tapered leading edge of the ducting
3) The internally tapered duct plus the flange

And here are the results - and they quite revealing.

Just tapering the internal leading edge of the duct adds an average of 6% more air speed across the 300 mm distance measured while the Bell mouth flange adds about 20%!

So overall well worth the effort.

Now I gotta build/find me one for the 6" duct opening.

good stuff

Mic-d certainly not my intention to dismiss published findings, and perhaps “theoretical” was a poor choice of words… but it is only when an idea / theory / application (call it what you will) is put into practice that it when it shows its true merit. As Bob said if it was mounted up high in the shed and was used to push air born invisible dust towards the extraction point with a clear line of sight, then I think this idea would be right in its element, and that would be an excellent application of it, but I would be extremely sceptical about its application where there is an any objects involved to deflect it. As we all know air that is pulled around and object acts completely different to air that is pushed around the same object from the opposite side.

BobL I did the exact same thing yesterday, went to Bunning and picked up a 4” floor flange, it caught my eye and I picked it up to see if I could use it for anything, but don’t bother reinventing the wheel, someone has already made the perfect 4” dust collection flange, they just call it a speaker port. You might even find some 6” ones out there, they get quite big.

http://www.simplyspeakers.com/speake...c-pt-f440.html

And here it is http://www.madisoundspeakerstore.com...lare-psp-6ofn/

Originally Posted by doug3030

OK, another thought comes to mind here about the graph at post #21, the speed did go up for the 4" intake on the later post, but why was the speed for the 4" lower than the speed for the 6" on that same graph?

Bob I know you took the 4" measurements later after I suggested it but did anything else change? I was thinking it may be due to turbulence as a result of a short length of 4" inserted into an otherwise 6" system but i didnt think that would make that much difference. Thoughts?

I haven't considered turbulence but now that I think about it the transition from the 4" to 6" diam duct is not smooth and quite short, is likely to be a contributor to variable air speeds at the opne duct.

The flow into the open duct is clearly turbulent since the needle on the velocity meter jumps around quite a bit (+/- 10% or more) during the measurements and in most cases (even the low speed readings) I have to eyeball an "average" reading. I am surprised the graphs turn out as smooth as they do. I have another meter that has a digital output to two decimal places and a smoothing function that I can use once I find it

Originally Posted by HeadScratcher

URL]http://www.simplyspeakers.com/speaker-port-tube-plastic-pt-f440.html[/URL] Madisound Speaker Store

Thanks for the lead. We have an audiophile at work that might have one I can borrow to try out. Another possibility is to turn a custom jobbie from some MDF. That way I can make it the right length and with a custom coupling to better fit the duct work.

RIght. That's it. I'm running a 4" line from the vacuum cleaner outlet to the nearest down-drop of the DC main. What about just a loose hood with airflow around the top of the vac to stop blower overload on the DC?

Originally Posted by veloaficionado

RIght. That's it. I'm running a 4" line from the vacuum cleaner outlet to the nearest down-drop of the DC main. What about just a loose hood with airflow around the top of the vac to stop blower overload on the DC?

I'm not quite sure what you mean by a loose hood but any air that comes out of a cheap vacuum cleaner and vented outside the shed is better than nothing.

I did a bit of a search for loudspeaker flared port profiles (I guessed that this would have been studied in some detail) and found this very interesting publication.
http://www.aes.org/tmpFiles/elib/20120821/11094.pdf

The study is for sound (ie two way flow) so the results have to be interpretted for one way air flow. For a simple flared inlet they split the flare profile up into
1) an edge radius
and
2) a longer taper.

The edge radius aspect is fairly straightforward.
An edge radius (r) to port diameter (D) r/D >0.2 produces a near lossless opening.
Thus for a 100 mm port the radius should be >20 mm, and >30 mm for a 150 mm port

Using a radiused opening at both ends of a tube supposedly produces up to 60% improvement in effective collecting area of the port compared to using not radiused entries. As only one end of our tube is radiused I'm guessing a 30% improvement may be possible.

The longer taper study is less applicable since it is studied by measuring sound distortion which may be linked to reducing turbulence.
A range of tapers were studied

s is the straight tube and sr is the simple 0.2 r/D radiused entry end.
There is some evidence that a longer taper reduces distortion over a sr but it's not that significant. One problem with using longer tapers is as you can see in the figure above they need to be relatively long ( for a 100 mm inlet the flare would need to be 200 mm long) and making them and being able to fit them into the spaces in and around ww machinery becomes problematic.

ANother other interesting study and conclusion is that roughening the surface to simulate a dimpled golf ball effect is counter productive.

Bob I don’t think everything will be directly relatable as a port is going to function significantly different to a constant vacuum source.
From my previous research one of the primary benefits is reduced pulsing / back pressure in the port due to the smoother transition to open air both sides, but this should still transition into a benefit with a constant vacuum source like a DC. Anything that helps guide the air into the tube has to help improve volumetric efficiency.