Thread: DE Measurements

Looks pretty good to me.

It's max speed is 25 m/s so you won't be able to directly measure the higher air speeds in the very middle of 4 and 6" ducting.
Don't be too worried about this as there are very few that can and they are VERY expensive
If you look at the sticky DRAFT: FAQ - Dust Extraction (Practical Aspects) it will show you how to use larger test ducts to perform these measurements.

It will measure up to 40 m/s, not 25 m/s. When I went over the calcs with the rep, 1200CFM in 6" pipe is approx, 6110.75 ft/min. and this will measure up to 7874 ft/min which is why he recommended this one.

Originally Posted by Lappa

It will measure up to 40 m/s, not 25 m/s. When I went over the calcs with the rep, 1200CFM in 6" pipe is approx, 6110.75 ft/min. and this will measure up to 7874 ft/min which is why he recommended this one.

6111 ft/s is a nominal speed . The air speed near the walls will be much less while in the middle of the duct will be at up to 1.4x higher so more like 8500 ft/min.

I just realised it won't matter much because the only time you'll see that sort of flow in a 6"duct with a naked impeller.
As soon as a few junctions and a machine is added the flow usually drops drop below 1200 CFM.

There's another really good reason for using a larger test duct though and that is so that air flow has some sort of a chance to stabilise before and after junctions etc. Otherwise the air flow drifts around over a time period that is long enough to drive you nuts making it really hard to even make a measurement. You might have to use the averaging feature of the unit to make sense of this. If you have to use a test duct you are better off using a bigger one and then the flow will be less turbulent making it much easier to measure.

Looks like a good unit. 40m/s according to the spec sheet which is more than high enough for a ducted system. The only addition that would be handy is a flange that can be attached to the pipe and that holds the probe steady in the position and depth you are measuring. Hopefully the probe has graduated markings along its length.

In my (relatively limited) experience, apart from at the very walls of the duct the velocity profile is fairly flat. That is, there isn't a huge difference between peak velocity at the centre of the duct vs the average. I found the same for my system and for a large workshop setup with a variety of 12" 10" 8" 6" and 4" ducting. The velocity profile was fairly flat.

Spent the past week boning up on airflow measurement with some people who do it for a living, the Internet and uni lectures and some very helpful tech advisers.
Problem No 1 was the manufacturers of the Hotwire I was interested in revised the spec by fitting a new prope in Dec. but didn't bother telling the OZ distributors so the OZ site showed 40m/s (that's what the OZ company went off) but the US site showed 25m/s so the top end was too low.
Problem No 2 is the cost of a Hotwire element and they are fragile
The general consensus was a pitot tube and some measuring device be it a liquid manometer, a magnahelic differential pressure gauge or a digital gauge.
I have some prices but I'm not sure I want to spend that much just for some home testing.
I was encouraged to make my own unit - although not accurate re actual figures, it would be accurate enough to show even minor variations as I made changes to my system. Pitot tubes you buy come with a K calibration factor and to get my crude device a calibration factor would have cost more that a real pitot tube.

I made one to suit my Super Vac system that consists of a Super Vac VC, a "$25" cyclone and 6M of flexy pvc pipe.

First I made a differential measuring tube consisting of a static tube and a pitot tube made from clear pvc tubing. It's crude but it works. I tried other set ups such as angle cut tube but this was by far the best.

pitot tube.JPG IMG_0432.jpg looking down the tube

I carried out 3 tests with the filter and the outlet pipe still connected to the Supervac.

Test 1. reading straight at inlet to Supervac VC that connectes to cyclone - 800 mm hose

before cyclone.jpg
before cyclone measurement.jpg Shows 82mm difference

test 2. reading straight after cyclone

after cyclcone.jpg
drop after cyclone.jpg see the difference
after cyclone measurement.jpg Shows 70mm difference

Test 3. At the end of the line with cyclone and 6M of hose

end of line.jpg
end of line measuremnent.jpg Show 61mm

Just interesting to see the drops as you go down the line.,

Now IF the device was accurate, which its not, calculations would have shown Test 1 - 115.28 CFM, Test 2 - 106.513 CFM and Test 3 - 99.43 CFM. The specs for the Super Vac are 180 CFM which, even with these inaccurate calcs, show is a stretch of the imagination.

Bugger about the anemometer but that is not the end of the world

25 m/s is still plenty high enough if you use a test duct larger than the duct you are testing which you really should use anyway. I do most of my measurements with a 20 m/s max. I do have a 30m/s anemometer but its not as reliable as the 20 m/s unit.

although not accurate re actual figures, it would be accurate enough to show even minor variations as I made changes to my system.

Even if you ignore accuracy and focus on reproducibility, if you don't use a test duct you will be really battling to see any real differences in most variations let alone minor ones. Minor variations take a long time to establish and can rarely be proven with single measurements. This is why you will see some of my posts reporting a value and then later on reporting a new value after I have performed more repeat measurements.

Originally Posted by Lappa

First I made a differential measuring tube consisting of a static tube and a pitot tube made from pvc pipe. It's crude but it works. I tried other set up such as angle cut tube but this was by far the best.

What size is the pipe?

If you look at the pictures, I am using a dedicated test pipe. I take that same test pipe from test point to test to test point hence the reliability of the readings. My main aim of the tests was to see the effect of having a cyclone in the system and the effect of the long flexy pipe. It sure showed the effect of those. I did each reading 4 times, in a sequence and achieved the same results.

The pipe size is DN40 as I was only testing my Vacuum Cleaner system as I stated in my post.

When I get around to testing the Dusty system, I will have a dedicated test pipe constructed in 8" pipe.

Sorry the pictures were not showing when I first looked at your post

Firstly good on you for having a go and I urge you to keep going.

Originally Posted by Lappa

Test 1. reading straight at inlet to Supervac VC that connectes to cyclone - 800 mm hose Shows 82mm difference

test 2. reading straight after cyclone Shows 70mm difference

That's a pressure loss of 12/82 *100 = 16%
That's considerably more than I would expect for a small cyclone connected to a shop vac, the Dust Deputy I tested lost between 5% to 7%
However that was for a 75 CFM Ryobi shop vac. Higher air flow rates VCs may lose more flow.

Test 3. At the end of the line with cyclone and 6M of hose Show 61mm
Just interesting to see the drops as you go down the line.,

Thats 9/70 *100 = 14%

And guess what going from an 800 mm long to a 6 m long 50 mm duct on a normally operating VC the pressure loss should be about 14%! BANG ON.
You can recheck this by removing the cyclone and testing a couple of different lengths of pipe.

What is the static pressure only of your VC - you might need a longer manometer (30+") to do this

Now IF the device was accurate, which its not, calculations would have shown Test 1 - 115.28 CFM, Test 2 - 106.513 CFM and Test 3 - 99.43 CFM. The specs for the Super Vac are 180 CFM which, even with these inaccurate calcs, show is a stretch of the imagination.

How are you calculating these air flow rates?
The flow Specs for the vac would be for no hose or filter and measuring the air speed in the middle of the duct, so ~100 CFM is about all I would expect.

I raise the following because if you do run into something that doesn't make sense this might help you understand why.
These effects have implications not just for absolute measurements but also for relative ones which is what you are trying to do
The Pitot tube setup you are using will be fine for relative measurements of slow fluid flow rates but at the air speeds involved with VCs the air flow will be turbulent and the pressure will fluctuate.
The pitot tube itself produces turbulence and the degree to which this effects the pressures may be non linear.

The fluctuations may occur both in space and time, and on short and sometimes longer timescales, and that is also dependent on flow rate.
The net effect is to create a system of small highs and lows within the pipe and this sets up a preferred air stream pathway within the pipe for a given flow rate.
Changing the flow rate by say a small modification to an inlet can shift the flow rate by a few % up or down and this changes the whole pattern of highs and lows in the pipe and moves the preferred air streams around inside the pipe.
You might register a increase in pressure but all that has happened is the pitot tube tips are now sitting in regions of greater pressure differential. You might calculate an increase in air speed but on average the air speed has gone down.

To reduce these effects a much larger (X10) test duct diameter than the diameter of the pitot tube is needed.
So if the pitot tube has an OD of 10 mm them a minimum of 100mm ducting should be used for the test duct
The test duct should also be ~10X longer than the diameter of the ducting being tested. If the test ducting diameter is larger than required the length of the test duct can be shorter

Calculating the air flow is done with an app from Dwyer, that the Dwyer tech rep. said to download. It has numerous fields of data you have to enter to arrive at the air flow rate. Its free to download for both Apple and Android devices. They also supplied me some literature - very helpful people and this type of testing is their bread and butter.

The 180 CFM for the Super Vac vacuum cleaner is what they claim on their site.

I understand the relationship between pitot tune diam and pipe diam and pipe length VS pipe ID. This device and testing was a bit of a "giggle" procedure from bits and pieces to see if it could be done simply and would variations show. I'm happy with the process. The urge to purchase a proper pitot tube and magnahelic gauge or digital meter is becoming stronger

Super Vac state 1522mm WC for static pressure.
I bought a long length of 3mm ID plastic tubing so I will put together a higher measuring manometer. Get back to you with the figures later.

Originally Posted by Lappa

Calculating the air flow is done with an app from Dwyer, that the Dwyer tech rep. said to download. It has numerous fields of data you have to enter to arrive at the air flow rate. Its free to download for both Apple and Android devices. They also supplied me some literature - very helpful people and this type of testing is their bread and butter.

We had Dwyer gear on the fume hood cabinets and clean rooms at work. Good stuff. I will download the app and knock up a giggle box and see what it does.

Super Vac state 1522mm WC for static pressure.

Well that's quite a bit - it should pull more than 100 CFM - maybe there's be a constriction inside of some kind.
To accurately measure the pressure difference the two ends of the Pitot tubes have to be at the same radial distance from the wall of the ducting. By measuring the pressure difference between the wall of the duct and the centre of the duct this will give an over estimate of the air speed so maybe the real speeds are not even

Should still give relative readings on larger differences.

I bought a long length of 3mm ID plastic tubing so I will put together a higher measuring manometer. Get back to you with the figures later.

To accurately measure the pressure difference the two ends of the Pitot tubes have to be at the same radial distance from the wall of the ducting. By measuring the pressure difference between the wall and the centre of the duct this will give an over estimate of the air speed

Thats interesting.

With the "pitot" in the centre of the air tube and the static tube level with the wall, as in test 3. above, I got 61mm difference, same as I got yesterday.

With the Static raised from the wall to the same height as the "pitot" tube (same radial distance ), I got 90mm difference.?

That's actually an increase air speed measurement - 7610.88 ft/min up from 6308.03 ft/min

You said it should go down?

Whoops wasn't thinking it through right. The air speed in the middle will be faster so bernoulli pressure will be lower so the P difference will be greater.

If the static tube is in front of the dynamic one there's also a chance the static tube is interfering with the air flow through to the dynamic tube. This could be tested by having a second static test point after the dynamic one although then the dynamic one may then interfere with the static. This is why pitot tubes are made like they are.

What it shows is the effect of the change in air speed with radial distance.
Now try moving the other tube down to the wall
The dynamic tube should be as close to the wall as possible , ideally within 5% of the duct diameter from the wall.

The static pipe height (in the new test) is level with the bottom of the pitot tube, its not blocking it.

All diagrams I have show the pitot tube (tube with front facing orifice) in the centre of the air stream and the vertical static tube level with the wall. That's how mine is set up in the first series of tests

Originally Posted by Lappa

The static pipe height (in the new test) is level with the bottom of the pitot tube, its not blocking it.

It doesn't have to be blocking the dynamic pipe directly to interfere with it - the end of the static pipe just below the level of the dynamic pipe can set up a disturbance than is significantly larger than itself and that perpetrates some distance down the pipe (see A below). This can either direct more or less air into the dynamic pipe. This problem is exacerbated by any sharp or ragged edges of the tubes.
If a long large test duct is used the two pipes can be separated apart so that the interference effect is substantially reduced.


Pitotpositions.jpg
Setup B above is what I meant in the second half of my previous post. This is just to check the opposite of what you are doing.
The average of the two measurements would give and indication of the average speed in the duct with minimal interference of the pipes on each other.
Now in practice you don't want to be swapping pipes back and forth for every measurement because you cannot be sure of replacing the tubes in the same place.
You could setup two sets of tubes permanently setup in a test pipe and swap the manometer or build another but that's also probably a bit much.
A couple of measurements of B and your original setup at different air speeds, would give you a set of fudge factors that you could apply to any setup.

Despite its problems I would use setup A (with the tubings spaced adequately apart) as it gives you the max pressure differential (max air speed) and is hence the most sensitive to changes.
Applying the fudge factors from above would give you a nominal indication of the average air speeds and hence flow

For those that don't know, pitot tubes are made so that the two pipes are incorporated into the one device (see photo below) and have smooth rounded edges which is one reason they are expensive.
Note the rounded end and the smooth edges of the static pressure holes at B
However, there are still interference issues with the static pressure measured at B, by air being disturbed by the interaction with the front of the tube at A - this why they have to be calibrated.
PitotTube.jpg

Last year I purchased some narrow brass tubing to have a go at making a small Pitot tube, but that project is sitting at the bottom of the todo list.
Getting one fine bent tube sitting inside another is rather tricky which is why I was going to use the setup as shown in the photo above.
Then I would calibrate the home made pitot against my hot wire anemometers and the big pitot tube, as shown in the picture above.