Thread: Improving machine cabinet dust ports

Originally Posted by Jeff Leslie

I intend running from a 200mm main line to 150mm steel with a minimum of 150 mm flex. Isn't that enough? Otherwise, I will have to run 200mm down, then choke it down via a tapered reducer to 150mm.

For a single machine duct that would be enough but it depends on the machine.

EG Up to a 300mm planer / thicknesser that would be enough

If its a 500 mm wide drum sander you might want to run an 200 mm down to the machine and/or split that off as 2 x 150 mm ducts to that machine.

Same for a TS which can use a 150 mm on the cabinet, and if can do it a 150 on the overhead guard.

You have to work out the economics of runs more 200 mm ducting on one gate down to a machine versus running 2 x 150 mm with two gates.

If you want detailed/specific advice you should post a diagram of your layout and machine specs.

Here's a ~$7 PVC fitting that some may not be aware of that can easily be converted into an approximate BMH for connection of ducting and flexy to machinery cabinets.

Its called a "Bolted endcap" - unfortunately I can only find it in 100 mm
Boltedendcap.jpg

Those SS screws can be undone an this is what is underneath
Boltedendcap2.jpg

By inserting a piece of 100 mm ducting into the inside as far as it can go and fixing it in place it is then possible to run a 16mm round-over bit around the inside - make sure the bearing can contact the PVC pipe.
The outermost lip of the flange can be removed using a BS or sanded way for a flush fit to a machine cabinet

Here is what it finally looks like compared to the original
BoltedendcapBMH.jpg

It's not ideal from a number of points of view.
The round over for 100 mm ducting should be a 25mm
The step in the flange should not be there but it cannot be removes but it could even be slightly rounded over.
However, even as is it will be considerably better than a hard 90º edge connect and the wide flange makes for a good strong connection to a cabinet.

I've actually been thinking of making a 150mm bell mouth out of timber and mounting it on a stand, would you say a 1 1/2" round over bit would be appropriate?

Originally Posted by bueller

I've actually been thinking of making a 150mm bell mouth out of timber and mounting it on a stand, would you say a 1 1/2" round over bit would be appropriate?

The optimum radius of round-over is 1/2 the radius of the duct so for 150 mm that is 37.5mm and given that 1.5" is 38.1mm I'd say that is pretty near optimised.

Beauty, cheers mate.

I was going to setup a different thread for this but it is so closely related to previous posts in this thread I thoughI had better add it to this one.

At the Mens shed there will be a need for quite a number of BMHs so I thought I'd see if I could find a way of making them cheaper than the previous posts and quicker than the MDF sandwich method.

The method is partially based on a Youtube method posted by Chris Parks ,so thanks to Chris for posting that.
It has taken me a few goes to get it working and further down I will show some failures as they are quite diagnostic.

The method uses standard PVC ducting and a wooden former mounted on an MDF base plate on a lathe like this

former.jpg
The former is made from just a crappy piece of green 250 mm diameter red gum branch I picked up at the tree loppers yard - hence the cracks.
I thought this would do as a trial but it has ended up working so well I won't bother about making a fancier one for plain 100 mm BMHs.

The former itself has a diameter of 100 mm at the narrow end (i.e. slightly smaller than the 102 mm ID of PVC ducting), while where the small step is in the former, right at the end of the flare, that's ~150 mm in diameter. Having a step turns out to be useful as described below.

The 150 mm step marks the final width at the end of the flare and is "duct diameter + half duct diameter" is ~150 mm which allows for a 25 mm of roundover all the way around.
The actual profile along the edge is not a simple "R + R/2" but somewhere between and R + R/2 and a bit of a more flared trumpet shape with the profile starting to flare away from the ~100 mm PVC diameter at about 3R from the end of the flare - I did this for no other reason because it's possible to do this using this method.

Using this method a number of profiles could be made - some could be short and stubby and follow the R + R/2 to fit inside cramped machine cabinets.

Below shows the setup with the BMH near completion.
A piece of PVC ducting is cut and the ends are sanded smooth and square.
The end of the duct being pushed onto the former has its inside edge relieved using half round file and sand paper and the inside of the tube is coated with light mineral oil/
The tailstock has a live centre and pushes on a stiff metal (Al) plate which pushes on the outer end of the duct. The plate could be made of wood.

A slight taper in the former enables the piece of ducting to be held semi-loosely in place and centred by tapping until it spins truly. It does not need to spin truly because it will be turned, but because you want the duct to be pushed on evenly.

The lathe is then run on slow speed (~60 RPM) [EDIT more like 120 rpm since I got my tacho for my lathe] while a hot air gun is played on the side and end of the duct on the former
I used the 350ºC setting on my hot air gun but I have no ideal how accurate that temperature setting is.
The process is speeded up if the former itself is preheated for at least 2 minutes and has mineral oil applied.
Also heat the exposed section of the former while making the flare.

If a strict R + R/2 profile is used the process could be done with out adjusting the position of the tailstock but in my case I had to reposition the tail stock to get enough horizontal movement.

setup.jpg

Once the BMH is formed there is a temptation to remove it immediately but it's best to let it cool on the former otherwise while it may deform/shrink.
You could use a wet cloth or spray gun but don't like the idea of water everywhere, so I just speed up the lathe a bit and play a now cool hot air gun across the end.
[EDIT I now use a damp cloth to help cool it down]
process2.jpg

Cooling the BMH on the former can mean it will shrink onto the former and become hard to remove.
This is where the step in the former comes in useful - while tapping firmly with a wooden mallet all over the outside of the BMH, apply a bit of leverage at various pints around the step with a wide blade screw drive it will eventually loosen.

And here is what it looks like.
BMHs1.jpg

Here is one of the failures shown on the right.
It appears I have overheated the PVC at a particular circumference and when I pushed the tailstock in, the PVC rucked as shown
BMHs.jpg

This was the first one I made.
I wanted to see how much I could flare the end and that is obviously too far.
I also did not have the hot air gun as hot on this one
Fail.jpg

Next step, make a former for the 6" BMH.

WE know this is good

bellmouth 3.jpg


WE know this is bad

bellmouth 1.jpg


Is this acceptable?

bellmouth 2.jpg

Cheers

Just to be clear - are these images of a right angle bend of a large blue duct attached to a smaller duct - or is it something else?

Also which way is the flow?

Its the LH side of an outlet from a machine made from a wood disk and a PCV pipe

bellmouth whole.JPG

Originally Posted by Lappa

Its the LH side of an outlet from a machine made from a wood disk and a PCV pipe

bellmouth whole.JPG

OK thanks.

So i n the diagram above it looks like there are 3 x sharp 90º bends all up, and they all will be, to some extent or other, resistant to flow.

One way to assess the resistance is to look at streamlines which indicate direction of fluid flow through/around a pipe/orifice/obstacle.

For a system with smooth transitions under very low pressures and flows, the streamlines will be evenly spaced apart and not crossing over. This is the lowest resistance situation and the flow is called non-turbulent.

As the flow rate increases, relative to the previous case the number of streamlines increases and the streamlines will be closer together in the middle of the duct and further apart near the wall.

At corners and transition this is where the streamlines will come very close together and if these transitions are not smooth, as soon as the stream lines notionally cross over this is when all hell breaks loose and turbulence happens and causes resistance. Stream lines are not supposed to cross over but as soon as they appear to cross is where

Without doing some fairly serious fluid dynamic modelling it is difficult to say exactly what the streamlines will when but, after looking at enough streamline models once can predict what most simple transitions will be like.

Here is my estimate of what that transition will do to streamlines at DC flow rates.

Each 90º corner will generate a small amount of turbulence.
A will probably generate the most because it represents a larger surface area of step.
B represents a small vacuum point in the flow so will also make some turbulence.
Can't really predict in this case (below) if B will make more turbulence than C - probably not.
Lappa3.jpg

So getting back to the original question "is this acceptable"?
While some rounding, even of individual components, is always better than none, none of the options showed in the diagrams are really close to optimal because the rounding over radii is so far away from the R/2 factor of the PVC pipe.
Rounding over the PVC pipe edge using say a 3mm radius won't do much and neither will the small radii round over applied to the wood disc.

The way to get closer to optimal is to round over the wood disk and the PVC pipe as a single entity like this.
Of course this often means that in your situation the PVC pipe will have very little wood to hang onto which is why a additional supporting piece (shown in Purple below) is sometimes needed.

Lappa4.jpg

And of course sometimes none of this is possible so as I have said before, you go for what you can get.

There is also poor mans round over (i.e. a simple taper to remove all hard 90º angles) like this.
It also leaves the same amount of PVC in contact with the wood disk.
Lappa5.jpg

The reason I asked was after serving the shape in your bell mouth on the lathe (looks like my third diag) and looking at the best alternative without going to 1/2r
Cheers

Originally Posted by Lappa

The reason I asked was after serving the shape in your bell mouth on the lathe (looks like my third diag) and looking at the best alternative without going to 1/2r
Cheers

The cross section dimensions of that hood are shown below.
The yellow MDF hood has a recess turned in the back and sits on the end of a slightly unusual 45º elbow (inner ID = 160mm), the inside lip of the elbow has a taper about 7 mm long.
In the photo your refer to the BMH has slightly slipped out of its recess.
I need to loosen a couple of screws at the back of the hood and refit it back onto the recess.
BMHLathe.jpg

Seems to me that you could sell these 5" and 6" hoods.

^ WHS

I'd buy a few of them at least.