Is there a better mousetrap? CFM saving designs? Changing direction costs big air!

[Bruce Seidner]

Read it and weep: http://www.engineeringtoolbox.com/minor-loss-air-ducts-fittings-d_208.html

I think I understand this but would appreciate any corrections. They seem to line up in general with the chart I found on Pentz site and the Air Handling System website.

Basically 90 degree duct joints cost lots of CFM or the equivalent straight pipe of 90 degrees (1.5 R/D) = 2.24 D, 90 degrees (2.0 R/D) =1.58 D, and
90 degrees (2.5 R/D) = 1.21 D

A 45 degree WYE=1.55 D.

So if I understand this a 6" pipe turned on a 90 degree radius of 1.5 (R/D) adds the equivalent of 13.44 feet of straight 6" duct to the system. Yowza!

There is less friction in a wider radius and so using the same 6" diameter duct on a radius of 2.5 is just adding 7.26 feet of additional duct.

Furthermore, am I correct that a 1.5 R/D for a 6" duct would be 9" tall and a 2.5 R/D would be 15" tall?

Is there another way to make a 90 degree turn, say combining shorter angle transitions with a short lengths of straight duct to achieve the 90 degree direction change with less height or does it matter?

I have to connect the top of a cyclone to a blower that will at best be at right angles and level to the inlet. At worst it will be 90 and a few feet below the level of the inlet.215551

[Jamie Buxton]

...Is there another way to make a 90 degree turn, say combining shorter angle transitions with a short lengths of straight duct to achieve the 90 degree direction change with less height or does it matter?...

Yes, bends in your ducts cost cfm. The larger you can make the radius, the less loss there will be. If you want, you can make the bends with short straight sections.

[Bruce Seidner]

Do the straight sections allow some speeding up before the next bend or are they just the removal of the friction fraught bends?

I wonder if anyone has put some thought into this compromise and recommended the most effective 90 degree joint with the least headroom?

I am looking into it but would prefer to not reinvent the wheel. It is not going to usher in an era of world peace but it is a curious puzzle of a problem that is interesting. Full of compromise, full of promise. All this is in a curve of PVC. Who would have thunk it?

[Dan Hintz]

Is there another way to make a 90 degree turn, say combining shorter angle transitions with a short lengths of straight duct to achieve the 90 degree direction change with less height or does it matter?
I use two 45s with a couple feet of straight in between...

[Lornie McCullough]

If it is only attic space above the cyclone, then there is nothing sacrosanct about cutting into it and mounting your blower between the rafters or above the rafters. You can always box around it.

If it is a second floor or basement situation, then.... well....... never mind.

Lornie

[Jamie Buxton]

Do the straight sections allow some speeding up before the next bend or are they just the removal of the friction fraught bends?

I wonder if anyone has put some thought into this compromise and recommended the most effective 90 degree joint with the least headroom?

I am looking into it but would prefer to not reinvent the wheel. It is not going to usher in an era of world peace but it is a curious puzzle of a problem that is interesting. Full of compromise, full of promise. All this is in a curve of PVC. Who would have thunk it?

Ah, PVC -- well you're on your own. My shop is plumbed with snap-lock HVAC ducting. In that stuff, there are adjustable bends. They're made of four or five straight sections whose ends are not parallel. Depending on how you twist each section with respect to the neighboring sections, you can get anything from 0 degree bends to 90 degrees. Stack several of them in series, and you can easily make a large-radius bend of any angle you choose.

[Anthony Whitesell]

Yes. I highly doubt my wife would like the cyclone motor mounted in a box protruding form the bedroom floor.

[David Kumm]

Industrial systems have the impeller and motor mounted away from the cyclone and although you need some extra ells you just have factor that into the size of the unit. Assuming you are pulling through rather than pushing into the cyclone you should use the same size pipe as the cyclone outlet which is generally larger than the mains. Exhausting into extra filter area will also compensate for the configuration. Dave

[David Giles]

A given blower has X available pressure drop. 90 degree turns burn up the pressure drop as do sudden contractions and expansions. For the same flow, a bigger diameter pipe (6") has way less pressure drop than a small diameter pipe (4"). Long sweeping turns are lower pressure drop than short sharp turns. But they take real estate. As you say, it is a tradeoff. In new system, a bigger blower could be added.

In any blower/duct work setup, the total pressure drop is the sum of all the individual pressure drops. In many cases, the pressure drop in one part of the system dominates the overall pressure drop. In many woodworking systems, the carefully engineered 6" ducts terminate with 4" machine ports. The pressure drop across the 4" port essentially sets the overall system pressure drop.

Just another thought to keep your mind spinning.

[ian maybury]

It's fairly typical Bruce to use tables which list an equivalent feet run of straight duct for bends when estimating total system pressure drop. You also need numbers for the blower and cyclone, the exhaust ducting and the filters - the number you would use on a fan curve to predict air flow is for the TOTAL run from intake at the machine hood, to exhausting from the filters.

This at whatever air speed you design for, and whatever ducting (there's some variation - PVC is apparently smooth and low resistance, spiral a little less). (the following are for the 4,000fpm recommended for good chip transportation up verticals) They can be found on the engineering data page of some ducting manufacturers.

Leakage can be a big factor on spiral ducting unless all of the joints are carefully taped. (up to perhaps 30 or 40% of the available CFM on a medium size shop run)

The figures I have (which I can't guarantee as accurate) have 100ft of straight 4in duct as creating 7in water gauge (WG) of pressure drop, or 100ft of 6in creating 4.5in WG

By that measure a standard 1.5 dia = rad 90 deg bend in 4in dia = 6ft run, in 6in dia = 12ft run. For a 45 deg bend 4in dia = 3ft run, 6in dia = 6ft. The effect of bends is fairly marked, it doesn't take a lot of 90 deg bends to chalk up the same resistance as 100ft of straight pipe. I'm not sure by that how much advantage that two 45 deg bends gives - although there probably is some. Unfortunately I don't have numbers for long radius bends.

The bad news is that for pressures below about 25in WG air is regarded as effectively incompressible, so you don't recover the pressure lost due to a local constriction like say a restrictive machine hood, or a tight bend.

This is also the reason why reducing branch sizes in drops to machines isn't a good idea, it just creates restriction

By this rule the air at our pressures behaves much like water in a pipe. The velocity remains constant from one end of the system to the other for a given pressure drop (fan suction) presuming that the cross section is constant. Against that the air will speed up going through a reduction, but it'll slow down again as the duct returns to the original size.

ian

[Jim O'Dell]

What I tried to follow in my design was to keep the runs as short as possible with the fewest turns. I used 2 45's with as long of a piece of straight pipe between them as I could per 90 degree turn in every situation. I wanted to have the cyclone in one corner with the main trunk going to the opposite corner, but that just wouldn't work. So I run down the middle of the shop and split off to each side. One side feeds 3 outlets, the other splits again and each split runs 2 machines. Small shop, 20 X 24. If you've got a large cyclone, chances are, unless you are in a huge shop, you can do whatever you want with turns and still have CFM to spare. Jim.

[ian maybury]

I ran into similar issues Jim, in that in order to get the ducts buried between ceiling joints and preserve my 8ft and bit head height I had to accept some extra bends and run distance. It's a Bruce says the bends that add resistance really quickly - a few extra feet of straight duct has minimal effect.

Another scenario people seem often to run into is that of needing extra bends to get hooked up to the cyclone and blower. This is a double whammy, in that it's reckoned advisable to have a decent length of straight run on the cyclone inlet (can't remember the number, but it amounts to maybe 6ft or a bit more on a 6in duct) as the turbulence can upset its fine dust performance a bit. Poking the cyclone through the ceiling into the loft can help get the inlet duct lined up straight, get some of the noise out of sour shop, and leave good height below for the chip drum and a decent length of connecting hose.

One to watch out for is that even high CFM low pressure blowers (the sort we use) still only function right in the typical say 5 - 12in water gauge pressure range (or whatever the fan curve says), so their output will tend to fall off much above that. Against that a larger (e.g. the 16in CV) impeller run at the same rpm (the 3,450rpm given by a two pole motor) as say a 14in has proportionally higher tip speed, and so its curve will probably show somewhat better performance at the higher pressure end of the curve.

Dropping the rpm by e.g. switching to a 50Hz supply (as in Europe) for the same reason causes the the fan performance to be most reduced at the high pressure end of the curve - which was why I found myself deciding to go for a VFD drive....

ian

[Bruce Seidner]

I went to Wikipedia and Googled about a bit and it seems that VFD is an electronic device that regulates the speed of the motor using frequency rather than reduced voltage like a potentiometer controller would.

This may be incorrect and I welcome correcting.

Are you introducing this device to a single phase motor or do you use a 3Phase motor?

Regardless, are you seeking to limit the speed of the motor to some optimal rpm or are you seeking to reduce the cost of running your motor?

I would have assumed the designers of the blower optimized the design knowing the motor they were going to employ in the design. Is this not the case?

Is there a more optimal rpm that one would research for the blower/motor combination that you are using?




I ran into similar issues Jim, in that in order to get the ducts buried between ceiling joints and preserve my 8ft and bit head height I had to accept some extra bends and run distance. It's a Bruce says the bends that add resistance really quickly - a few extra feet of straight duct has minimal effect.

Another scenario people seem often to run into is that of needing extra bends to get hooked up to the cyclone and blower. This is a double whammy, in that it's reckoned advisable to have a decent length of straight run on the cyclone inlet (can't remember the number, but it amounts to maybe 6ft or a bit more on a 6in duct) as the turbulence can upset its fine dust performance a bit. Poking the cyclone through the ceiling into the loft can help get the inlet duct lined up straight, get some of the noise out of sour shop, and leave good height below for the chip drum and a decent length of connecting hose.

One to watch out for is that even high CFM low pressure blowers (the sort we use) still only function right in the typical say 5 - 12in water gauge pressure range (or whatever the fan curve says), so their output will tend to fall off much above that. Against that a larger (e.g. the 16in CV) impeller run at the same rpm (the 3,450rpm given by a two pole motor) as say a 14in has proportionally higher tip speed, and so its curve will probably show somewhat better performance at the higher pressure end of the curve.

Dropping the rpm by e.g. switching to a 50Hz supply (as in Europe) for the same reason causes the the fan performance to be most reduced at the high pressure end of the curve - which was why I found myself deciding to go for a VFD drive....

ian

[ian maybury]

A VFD Bruce is a device that depending on the unit takes in single or three phase power, and outputs three phase power - but at whatever frequency it's been set to deliver. They use electronic switching devices to build the voltage curve for each phase, so the output is in effect a synthesised (pulsed or digital) version of the analogue normal. They have become very sophisticated, and are not so expensive as to be out of reach for the likes of ourselves.

Some capabilities/tasks they handle include:

- Running a three phase motor from a single phase supply. (but the output voltage is the same as the input, so over here that normally means 230V 3phase or a dual voltage motor - unless a transformer is used to step up to the higher 400V industrial 3 phase voltage which can bring some problems)
- Providing the ability to greatly vary the frequency of the output power, although motors typically only use up to about +/- 20% of the frequency of that input due to mechanical and cooling related limitations. e.g. changes 50 to 60Hz, or vice versa, or as set. Changing the frequency of the power fed to a motor changes its speed in proportion to the change in frequency. A VFD can speed up and slow down a motor to a pre-determined pattern in response to input signals too. (see below)
- Providing the ability to control voltage and current while starting a motor so that the peak current demanded from the supply during e.g. a 10 sec ramp up to speed can be as little as 140% of the steady state load amps. (a single phase motor may draw 5 or 6 times its full load current when starting)
- Providing the ability (via low voltage input terminals, and the associated built-in control circuits) to be remotely controlled using manual switches, radio operated relays (switches) or the output from a computer or PLC controller to start/stop/vary speed/reverse/brake to a stop etc.
- Providing lots of protection to the itself and the driven motor by monitoring lots of variables relating to power drawn and comparing these to what it figures is normal for such a load. This is all pretty much automatic, and at most requires keying in basic information about the driven motor.

Limitations include a tendency to generate a fair amount of electrical and EMF noise (cabling needs to be shielded, it's best to use a filter (manufacturer approved models are usually sold by the same supplier as a low cost ready to use box) on the power input etc, and they can sometimes cause problems with spurious trips if wired through an RCD (earth leakage device) protected circuit), they ideally should for safety be mounted in a well ventilated electrical enclosure (they need some cooling air), and need typically to be connected to a single load only. i.e. you can run multiple motors at once, but only if they are wired together from the VFD output. It's more typical to use them to drive one motor only, and that needs to be connected directly which means that it often involves by-passing the controls on machines, and controlling them directly from the VFD. (the various switching that e.g. a star delta starter as might be found in many three phase motor installations goes through will confuse the brains of the VFD)

I'm no expert by the way, but installing my dust system required finding a way to run a 5hp (actually 4kW) motor (which in 230V/50Hz single phase form could if available draw up to 80A on start up) from a 63A 230V domestic supply without overloading it (the supply). Which with some much appreciated help from others here (and lots of reading and head scratching) led to the decision to use a VFD. The VFD added speed control which suited too as the Clear Vue blower is designed to run at the 3,450rpm you get with 60Hz power in the US, but which drops to 2,850 rpm on our 50Hz power. Also ease of remote control (no relays etc required) since the control circuits are 24v and draw only miliamps.

It's all running smoothly with no evident problems - so far anyway....

ian

[Bruce Seidner]

I have a 220v 5hp single phase motor running my blower. I can get the fan curve from the company and actually think I have it in a file somewhere or other on my computer.

What is it that you were wanting to achieve? My sense is that there is a point on the CFM curve related to the blower at different horse power that uses less electricity for the CFM you require. Do you change these VFD settings for different combinations of machines running at the same time or for different machines? I think I still fail to understand the reasons you would go to these lengths and what you would achieve that would explain the lengths to which you have gone. Not that anyone has to explain the primal instinct to acquire tools and gadgets, I get that, boy do I get that. What is the end result that you have sought?

[ian maybury]

It was nothing to do with tools and gadgets Bruce, but a necessity. I'd have been happy to save the probably $500 extra cost it added to my installation.

My two needs were (a) to step the speed of my European 4kW 50Hz motor (running on my irish 50Hz supply - US supplies are at 60Hz, and give 3,450RPM on the two pole motors normally used on dust systems) from 2,850rpm up to the 3,450rpm the Clear Vue blower is designed for, and (b) to control the current draw to something the service provider fitted 62A main fuse on my supply could tolerate while starting the motor. (you guys in the US seem to have much higher amps limits on your supplies, so starting a single phase or large motor is no problem for you)

The reason for (a) is that centrifugal fans lose performance mostly at the high pressure end of the fan curve when the rpm at which they are run is reduced. I might have been OK at 2,850 rpm, but wasn't quite sure what system pressure drop I was going to end up with, and felt it best to take no chances.

(b) I've just explained, it was what made the VFD essential if I was to be able to run the 4kW/5HP motor the full power Clear Vue system requires on my 62A limited irish supply. We can't for this reason even normally source a 5hp single phase motor over here, 2.2KW or 3hp is typically the largest sold.

You could certainly use a VFD to ramp the speed of your blower and hence your CFM and pressure up and down if you wanted to, but it would only be useful in certain circumstances. It might for example suit to do this to tune out a vibration or harmonic making noise, or you might have a machine where having the full suction or not enough causes problems.

This latter in effect seems to be what Oneida are doing with their 'smart' system. Chances are they restrict the flow and up the rpm to get higher pressure t lower CFM for vacuum use with small hoses, but use the VFD to reduce the RPM and hence the power drawn when running with a large duct at high CFM in normal low pressure dust collection applications.

The problem with using a VFD to reduce air volume (if you say had a system running multiple machines, and you wanted to reduce to say one) is that not only does reducing the RPM reduce the air volume, it also reduces suction/pressure - perhaps too low to work properly. (how much is determined by the change in the fan curves for the differing RPM)

The other reason it probably would not be so useful run that way is that it's not really needed - simply closing the blast gates on the lines to the machines not in use throttles the fan intake, and as a result drops the CFM and the power draw accordingly.

ian

[Bruce Seidner]

I am sorry if there was any potential reading of my question to suggest that it was gratuitous or mere gadgetry to use a VFD. I did not understand the reason for the VFD but clearly if you're native lines are 50HZ and the blower motor is expecting 60HZ to perform optimally you need to make the adjustments. It is also fascinating to learn that there is an alternate to buying a second motor of equal size to manage 3phase requirements for those of us in residential neighborhoods without any recourse to 3phase lines. Thank you for the tutorial. I have learned quite a lot.

It really is an injustice.

I can open a Guinness in the S.E. of North America and regardless of the latitude between us it works just fine.

I do wish our motors had the same tolerance.

[David Kumm]

Bruce, i like Ian run a DC system with a three phase motor and a vfd. I'm in the US but here are my reasons. The dust collector motor is the hardest working motor in the shop. The single phase Baldor 5hp motor runs at a FLA of 19 and is less efficient than the 12 amp 3 phase motor so in reality you are getting a smaller motor in single phase. There have been lots of discussion about how often to start and stop single phase motors due to the immediate load and whether frequent starts will cause premature failure. The three phase motor and the soft start capability of the vfd eliminate those problems and reduce the amp draw at start up. Although you can get a system curve from the manufacturer it will be a theoretical calculation based on a million assumptions that may or may not factual for what is finally built. Pipe size, bends, filter area, filter clogging, amount of flex used, on and on. When you get all done you might find that when you put an ampmeter on the motor that you are pulling too much or too few amps when various gates are open. My system has 21 gates ranging from 4" to 7". My 24" old Oliver planer and Oakley edge sander are at the end of the run, 50 ft away. Great planning on my part. To maximize the cfm at those ports which means drawing 12 amps, or slightly more means I can run the motor at 63 hz. That increases CFM by 12-15% which is a big deal with old machines not known for great dust collection. For the price of a used vfd I could get a bigger, better motor, fine tune my system, have built in overload protection, and a soft start capability. Basically for free. I set the vfd to display amps and can tell when the filters need to be changed by the reduction in amp draw when certain gates are open. I'm a believer that it is worth the effort once you get to the 5 hp level at least. Dave