Axial Flow Fan

Alas...if only it were so easy.

lolx. i know its not dat easy :-P

but does something like dat exist?

Velocicalc Air Velocity Meters | TSI Inc

You can refer any fan engineering handbooks such as Fan Application Guide published by the Heavc Association. Usually axial flow fans are used for handling large volume of air and be used where noise level considerations are not important. There is a general formula

Required Work by fan = Flow rate X Fan total pressure divided by fan efficiency. Fan characteristic curves are available with the manufacturers. Easy way is get the model form the name plate and refer the technical catalogue of the product.

The Formulae will only give you part of the answer. The required efficiency will be from the manufacturer test rigs.

Theoretically, you can work it to have an idea but really very crude: The pitch angle will give you the pitch length.This is the length of the cylinder of diametre = the fan effective diametre. This cylinder volume multiply by the rpm will give the volume moved per minute (all is very rough and simplified). The number of blades is to improve the pressure efficiency of the fan: that is the lowering of the back leaking of air, due to the resistance of the existing air at the front of the fan, from being moved forward. This combination will dictate the power required to keep the fan rotating at the required speed at the given conditions.

That is how far I can go without finding the formulae the manufacturer will be using (includes the different factors affecting the efficiency , air density etc...

IIRC that is the type of information covered in 'psychrometrics'.

Also,in ducted fan applications (similar if not the same),a magic number called 'swept area' comes into play.

Thats about all I can tell ya.

Skies.

Jay.

If you can find a copy of this book, it contains what you seek.

A further factor for you list is "solidity".

i have used a formula and compared the theoretical and actual flow rates. there is a lot of difference :-(

Show us the formula used (complete with symbol definitions) and your results - both math and measured.

And it would be handy to know who you are replying to - i.e. use "reply" to the poster in their post window, not the "reply" at the bottom of the thread.

i used the following formula for the air flow rate:

Flow Rate = Q = Vd*Nb*Nr*60

Q = flow rate in (m3/hr)

where Vd = volume of air displaced by single blade during one rotation. this is the volume i have taken as the volume equal to the volume of cylinder constructed by rotating that blade in 360 degree. ( in m3 )

Nb = No. of blades

Nr = Rotation speed in rpm

but i am getting a lot of difference in measured reading and calculated reading

Well for a start all the blades produce that singular Vd per rotation - so multiplying by n blades will give a error of 'n' times.

Secondly the pitch of the blades matters in terms of back pressure - as does 'solidity' or the amount of 'daylight' you can see though the fan.

What design of fan is it?

This is then fan front view, its blades are straight ( no rotation about center )

"Well for a start all the blades produce that singular Vd per rotation - so multiplying by n blades will give a error of 'n' times."

i don't understand from this " n " times error

regards

You got the length of the 'cylinder of air' by plotting the helix produced by the pitch of one blade - yes?

But all the blades are doing that same helix in that same cylinder of air at the same time. Each is moving it's part of the cylinder of air - or "illustratively" in the above, one blade would only move one eighth of the cylinder volume like a 1/8 area helix.

But; the cylinder volume you calculated is already "eight 1/8 helix's" - so multiplying again by 8 blades is a 'false multiplier' (like 64 out).

Clear?

It's a bit tricky, but all 'rotor discs' are assumed a 'solid mono blade' like a perfect Archimedes screw, for 'this theoretical volume part' of the exercise.

And later that 'solid' becomes 'blades and not blades', and "losses" rear their ugly heads.

if i am not mistaken: The volume of air displaced by one blade per rotation is equal to that cylinder volume. . so if there are 8 number of blades rotating, why shouldn't we multiply it with 8 to get the total displaced volume, because each blade is displacing air volume saperately. it has nothing to do with the preceeding blade. whenever a blade passes a point by displacing the air in that region, before the incoming blade, the air occupies that displaced volume too, so i guess that 8th time error is not obvious. please correct me if i am wrong.

i got the length of the cylinder by taking the average height of the blade using the pitch angle.

I did

I think that is the main misconception: One blade per rotation would not produce the full theoretical cylinder volume. It would only be about the same fraction of that volume as the fraction of the circle the vane's projected area occupies. In the picture, each blade occupies maybe 1/12 of the entire circle. Furthermore, the blades are most likely twisted so that the helix angle is steeper nearer the center to compensate for the lower speed of the portion of the blade near the center. (To say nothing of camber and other airfoil parameters.)

i have checked the pitch angle both from the tip and the base of the blade, and its about the same.

In that case, it may be a cheapo fan design with even less efficiency than previously thought.

For your Info:

The Pitch length λ = 3.1415 x 55.79/ Tan(pitch angle or what is called the helical angle).

if diameter is in cm then Length λ is in cm

The Volume of the cylinder to calculate: Vd = λ * (D^2 - d^2) x Π /4 (Π = 3.1415)

where d = 55.79 and D = 55.79 + 46.61

this gives Vd = λ x 5791 approx. (cm³ ??)

We need to know the Blade pitch angle to complete the Calculation for Vd = Volume swept per One Revolution.

the number of blades will only come to evaluate the efficiency: the more blades, the more the REAL volume swept in will be near the theoretical, therefore improving the pressure produced forward.

Q = total volume per minute, = Vd x rpm

If you have some mechacical engineering back ground, then, As a comparison to understand why the number of blades (n) is irrelevant for the total volume swept theoretically, immagine it to be a Worm Wheel with multiple starting threads on the same worm...the pitch length (the amount of travel per rev.) will be = to the pitch of one thread (helical)...(?)

the blade pitch angle is 17 degree and the rpm = 750

what would be the Vd value according to your calculation?

where d = 55.79 and D = 55.79 + 46.61

i think D = 55.79 + 2(46.61)

"i think D = 55.79 + 2(46.61)"

And you would be right

BTW are we in mm or cm?

(I gather we are not in inches from your m³ mention)

the dimensions are in cm :-)

suppose there is only one blade on the fan, and what would be volume of the cylinder per rotation then?

Ok - but cm confuse me horribly as they are a non-preferred metric unit, popular with for dress makers, not engineers.

Even-so - I can 'convert' - (but could screw up, but luckily Tornado math king, has turned up to pounce on my math bungles).

Couple of extra questions;

How did you measure the pitch?

Where did you measure it? - at the tip - at the blade root (hub) - or mid blade.

Are the tip and root width dimensions 'square' to the view, or square to the blade?

Are the blades 'twisted' or 'straight'?

Are the blades 'cambered' - i.e. have a 'C' shape section?

(they could be flat sheet, maybe twisted - or cambered, with or without twist - this changes, along with the taper dimensions, what momentum is imparted)

Is this fan in a ring?, or duct? or in free air?

(this relates to radial flow rates and 'tip losses')

Now the questions above - answered - will put you in the position to solve this properly.

I am assuming you have acceded in the face of 3 people telling you "n blades" is not applicable to your approach? Y/N?

If so; as you can see the 550 hub is not part of the fan disc either, and its volume has to be filled by the 'doughnut' the blades move.

So though the pitch x area describes the 'doughnut' length - it does not describe the volume in a duct.

I.e. your duct velocity will be reduced by the doughnut having to fill the center void.

i think most of the answers u can get from this picture,

How & where did u measure the pitch:

pitch was measured using clinometer both from tip and base of the blade, found almost equal.

Are the blades twisted or straight:

blades are straight

Are the blades Cambered:

No blades ain't cambered

Is this fan in Ring of Duct:

This fan is in a ring, two parallel fans are sucking air from rotary air filters.

and yes i admit that n blades have nothing to do here. but......

Well if this is your design / proposal / solution - you have a long way to go in understanding fans.

The reason trapezoidal blades of 'same camber' (curved section like one half of yours); is the chord length changes, so the pitch changes.

You have almost every concept back to front in this design.

Here are some pictures you should contemplate;

You are right, my mistake!

The total volume displaced per minute will be (Theoretically, not considering backflows, air resistance, and others...) = 6446.82 m3/minute, and 386809 m3/h

(I hope I did not mess up the calc.)

This is assuming the pitch angle is correct reasonably.

Of course, this is only theoretically, without any consideration for the efficiencies, losses, backflow , air resistance forward...

The blade twist is usually such that when the angle at the base is used for the pitch length calc, it should give the same pitch length with the calc using the angle at the tip (ideally), otherwise you will have losses and vibrations (some other turbulance... not being an expert in the field.... just from basic mechanical engineering view point.). Blade shape workout is quiet an expertise and needs a lot of unknown variables obtained in a wind tunnel testing....

Then the formula is codswallop - in the same avenue as the proverbial chocolate teapot and the one-piece book-end.

it would be appreciatable if you can suggest the correct formula then :-P

i have rotated the effective pitch of blades in solid edge and the resulting cylinder is as shown:

the volume of this cylinder is about 74596984.163 mm^3 as measured by the software

now if it is assumed that per rotation, the volume of air displaced is as mentioned above, then the fan with 750 rpm should produce flow rate as

Q = 0.0746*750*60 = 3357 m3/hr

which is way too small when i compare to the measured flow. so i guess the calculation must be wrong :-(

That picture doesn't look even remotely realistic, I'm afraid.

can u elaborate plz

Nar I think he's just being annoying

730 dia = 2293.4 circ

127 tip = 2293.4/127 = 18

tan 17 = .305

.305 x 127 x 18 = 698.76

I think

I'll start by assuming the fan is an idealized screw of one turn, with an unused center of 0.6m diameter, an o.d. of 1.56m, and an "average" mid-blade diameter of 1.08m. If the air flow were perfect, it would advance by l = π d_avg sin 17° per revolution. In this case, l ≈ 3.14 (1.08) (0.29) ≈ 0.98m per revolution. This length would be from upper left to lower right in the picture; thus your "washer" is way too thin.

This theoretical advance would be reduced by the solidity factor of the fan, which looks like about 0.5 from the "head-on" blade picture. It will be further reduced to the fan efficiency (backflow between blades and between the blade tips and the duct), which by wild guess could be 0.7. With all that, 0.98 x 0.5 x 0.7 ≈ 0.343m of advance per revolution.

For the annular region in question, the resulting volume is about (π/4) (1.56² - 0.60²) (0.343) ≈ 0.56m³ per revolution. Then x 750 rpm x 60 = 25,200 m³/h. Is this a bit closer to your measured flow?

(According to the most recent picture, the blades are cambered, but I didn't factor that in.)

Tornado, I worked it this way.

Now using the efficiency etc... (25200/386819==>6.5%)

My earlier reply said sin 17°; should have been tan 17°. Good catch for those who noticed.

I think pitch length should be diameter * pi * tan 17° (rather than ÷ tan 17°).

Oh no! I'm 2.39 mm wrong! ^{_{(if the diameter was 730, not 1490)}}

The Pitch Length Formula is diameter * pi /tan(Helical angle).

This is from the projection of the helical on the plane which gives a sinusoid, with an Amplitude = Hub/2.

The error is from the angle given: He gave the angle of the blade with the Fan diametre, while we need to have the angle with the axis of the fan (my mistake). This will be 90-17 = 73 degrees. The calulation will be:

Which is more correct. And your estimation will be with an efficiency of ~70%. I thing it might be less than that but I cannot say it without the unknown.

Here is the file containing the air flow from each fan.

http://www.mediafire.com/?h3dmx69csbbdjkd

Check out the measured air flow rates of the fans. these were measured at 780 RPM

Thanks for the wild goose chase. The link gave some pop-up crap, but no fan data.

You gota find the

in that mess and it gives you file that opens in excel.

oops. . sorry it links doesn't work for you people.

i don't know how to upload file in this forum :@

i have pasted the tables from XL file. it mixed the columns :@

sorry!

I couldn't find in the worksheet any fan whose area matched the dimensions discussed so far. Which of these fans is in question, anyway?

Besides that, the usual shtick on fan capacities is to get the tables from the factory, rather than trying to reverse-engineer them. (Unless you suspect they sent you one with the wrong blade angle, for instance.)

The idea was only to get u know about the actual flow rates of these fans. if i were having the data sheets of these fans from the manufecturer, it would have been so easy :-@

the area mentioned is not of the fan, instead the duct from which it is sucking the air.

From the data shown:

1- If the cross section of the suction duct is Uniform and the different measuring points are for the same cross section, then the average is acceptable. Otherwise you cannot average it.

2- If you select the one Fan that matches youe fan (i.e. the one you have the dimensions of the blades and the base angle for it), then you can use the formulae I gave in my last contribution to obtain the maximum theoretical flow and compare with the flow worked out in the table/measurement of the excel file. the ratio will give the efficiency at the given conditions when the measurements were made.

3- All this will still need corrections each time the media condition varies. that is why the Normal conditions of the testing and provuded data is allways stated when such table are given by the manufacturers...

Yes the duct's cross section is same and the readings were also taken form the same cross section.

The maximum flow rate calculated seems to be less than the measured flow rate, so it would make the efficiency greater than 100 :-@

The calculations depend greatly on the accuracy of the Helical angle measurement (pitch angle of the blade).

for your 17deg = 73deg pitch angle == 36,152 m3/h

for your 18deg = 72 deg pitch angle == 38,425 m3/h

for your 20deg = 70 deg pitch angle == 43,044 m3/h

Also, the accuracy of the air speed measuement: If the speed was measured at the same cross section point, but at different positions in the duct ( near the duct surface, near the centre... ) the average will not be a simple straight forward calculation as shown in the table. The flow has to be laminar without any turbulence etc...

The best way is to take the average static/dynamic pressures and work out the velocities.

you're right that the change in pitch angle have a great effect on the flow rates thats why i used clinometer to measure the pitch angles and it was exactly 17 degree.

however there can be issue with velocities because the ducts are of uniform cross section but there are certain hurdles in the ducts in the form of compressor pipe, electrical wirings and cotton heaps :@

i should start measuring the velocities using the static/dynamic pressure now to verify the results.

Also there are two fans with FRP blades. May be those fans have greater efficiencies producing more air flow than the normal ones

i wonder the dynamic pressure would also varry greatly at different points in the duct because of varrying air velocity ??

Yes, Measure Static and Total Pressure at different point: Centre, Top, Bottom, Left, Right and in between.

Average the results and then work out the dynamic pressure to calculate the average Velocity. In a square duct cross section, ignore the points at the corners.