What Kind of Fan Provides the Best Pressure / Airflow?

Fan laws can be used to determine the affect on fan performance

due to changing a parameter. There are several simple

relationships among fan capacity, pressure, speed and power.

These relationships are referred to as the fan laws.

· Capacity is directly proportional to fan speed,

Q µ rpm.

· Pressure is directly proportional to the square of the

fan speed, P µ rpm2.

· Power is directly proportional to the cube of the fan

speed, HP µ rpm3.

The above relationships apply to fans of fixed geometry.

When you can change the fan size the following

relationships at constant fan speed, wheel width and

pressure apply.

· Capacity is directly proportional to the wheel diameter

cubed, Q µ D3.

· Pressure is directly proportional to the square of the

wheel diameter, P µ D2.

· Power is directly proportional to the wheel diameter to

the fifth power, HP µ D5.

Your 47,100 mm³ should be divided by 10⁹ to convert to m³, so it should be 4.71 x 10^-5 m³. But actually this volume is not meaningful to your application. Your pressure calculation looked okay. I don't usually work in SI, but this pressure I think is about equal to 20 mm H₂O, which is the type of units used to rate the static pressure of fans. You need a fan that will operate at this pressure. I'm not sure what small axial fans might meet this, but centrifugals can do well over this. If I were just playing around with this, I would try an air bed inflator to start with; they are inexpensive.

What you really need to know is flow. I'm not expert ehough to calculate the requirement for this.

A DustBuster would work too...

In fact I did not read your comment with great attention previously, I have to make now a correction:

"actually this volume is not meaningful to your application."

This statement is wrong since as well the overpressure as the flow through the gap between platform and ground have to be satisfied at same time.

If we consider the gap of 2mm and a gap length of 550mm with a pressure difference at the skirt of 312.5 N/m² the flow which has to be delivered at the pressure of 312.5 N/m² is about 0.0145 m^3/s, this leads to a power requirement of 4.5 W net.

The efficiency of small motors is low so that electrically the power will have to be over 10W.

Sorry for the late reaction.

You are considering pressurizing the skirt and venting the skirt into the center enabling a 2mm gap between the skirt and the ground ??

Thanks for the speedy answers,

@ Tornado,

Thanks for the correction, 4.71 x 10^-5 m³ sounds a lot better, its been 27 yrs. since ive been in any type of class room and you were close according to my unit converter it would be 31.77mm H²O.

With my "try it before I think about it" mentality, I tried a travel hair dryer

first, which has volume but no pressure and that's why I need help with a different fan setup.

@bwire,

I am sure you have just given me the answer to my question on the dimensions of the fan required, but it will take a bit for my brain to put that information in the right order. So thanks.

Thanks for all the great information so far,

To let everyone know what im trying to accomplish, here are some pictures of what I was planning.

This pad will be skirted to act like a hovercraft.

The inlet is off centred to allow for the thrust and servo

This is just a subsystem of the complete vehicle which shouldn't weigh more than 3Kg.

I believe that the hair dryer will give me enough thrust ( 4 pads ) once I can get the pad off the surface.

Thanks for all input.

Amilo ( admin@russ-smith.de )

Just a thought, having the thrust off centre is also going to create a turning moment, the pad will want to rotate to the right. If you are using 4 pads, will you have 2 R.H & 2 L.H.

Hello Amilo

You need the average air pressure under the thing to = the 312 Pa you calculated. An interesting exercise to work out the pressure drop per mm distance from the centre, as the velocity falls the further out you go due to increasing flow area. The pressure drop lets you find the pressure profile hence average pressure. I'd assume circular rather than oval as it'll be hard enough without that extra complication.

I might have a go at it if I have a minute. But my guess is the pressure at the centre (that the blower must provide) is a lot higher than the average pressure. Probably easiest to assume an air flow and find the pressure, then adjust to suit.

Cheers...........Codey

At this pressure drop the flow can be near to laminar also because gap/length is small (no skirt). For this conditions the pressure drop is

Δp(r)=Δp(ri)*(1- [ln(r/ri)]/[ln(re/ri)] with r- a radius between ri and re

ri= internal radius of supporting area

re= external radius of supporting area

ln = natural logarithm

Δp(ri)= over pressure at the internal radius as in my sketch above.

For a ratio re/ri small the variation is almost linear, for bigger ratios the logarithmic function leads to a non linear pressure distribution which I have not in my sketch for reasons of simplicity.

Codemaster is right : with a round area the flow is completely symmetric, with an oval plate the flow will choose the minimal resistance, it will be higher in the direction of the small half diameter of ellipse. All above is valid WITHOUT skirt.

if you clip all the rectifier diodes out of a regular 110v a/c hair drier, you end up with the same fan you just showed a pic of, it runs off d/c as well. try your fan at a higher voltage say 18 volts. see if that gives you more pressure.It wont last as long, but may help you in your quest for the right fan. or try an r/c hobby shop for a ducted fan motor meant to move things.

I did clip the rectifier diodes from a 220v a/c hair drier and am running them from this power supply,

4 X 7.2v 3800mAh hooked up in series / parallel, at full charge, I have 15.6v.

To discharge the power supply, I hooked up all 8 motors and they ran for about 35 min. till the voltage dropped down to 12v after that it dropped rapidly.

I believe the problem lies in the fan blades,

The angle lies at 45°, is this too much of an angle, I need more pressure for lift. Plus there is a gap of about 1mm between the blade and the casing, will closing this gap help/give me more pressure?

Many thanks for all who have participated so far.

Amilo

Plus there is a gap of about 1mm between the blade and the casing, will closing this gap help/give me more pressure?

No

Don't know if you are interested, but as I recall it there could be advantages in using a skirt - as they do in hovercraft.

One advantage that this gives is ground clearance (except under the skirt) and perhaps of more relevance to you I seem to remember that directing the flow at the edge of the disc downward, and slightly inward.

This uses the dynamic head property of the jet stream to pressurise the area under the disc and so give much better lift for the energy input. Its a bit like the principle of a small stand pipe of water on the top of a bladder that pressurises the entire contents of the bladder (and causes huge bursting forces)

The kind of fan you have there is similar to computer "muffin" fans. Those I found on the Web were good for only about 8 mm H2O static pressure, vs the 31 you calculated (my guess of 20 was just that, a quick guess). There are multi-stage axial fans that can go higher, but I doubt they will be easily found in small sizes. Your earlier mention of centrifugal is likely to be better. I found my air bed fan, and if I can find the time I will cut a hole in a board and try it out (powered by four D cells, thus rather heavy). Your initial estimate was 750 g for the assembly; a later estimate was 3 kg, so the required fan pressure will increase accordingly.

This is a fun project, and there has been lots of good input. Best of luck, and please keep us posted.

Thanks Tornado,

The 750g was given as the one pads share of the whole, 4 pads plus body should weigh approx. 3Kg total.

Imagine my ATV with hover pads instead of wheels, shrunk down to RC size of course.

Amilo

I wondered about the business of the 4 pads; this clarifies it perfectly. Thanks.

If I can do it, I plan to locate the hole in the board toward the "front" of a single-pad tryout. The hole will be angled down and toward the "rear." The idea is to provide both lift and thrust with just one fan, which is all I have. If I had the tooling, I would make this as a slide so that I could maybe balance everything for a constant-height periphery all around. As it is, I will at first be making only a wild guess. This version won't be steerable, but hey, one or two things at a time. I can glue a strip of rubber around this, so it will be skirted, as in your plan.

If the volume were reduced to the right front pad and slightly of the right rear pad would it be enough to orient the quad towards the right and vice verse towards the left?

My plan was to implement RC servos into the A-Arms ( where they attach to the pads ) of the craft to twist the pads in the desired direction ( front and back ) to allow the thrust motors to steer the craft.

Amilo

Curious, what other mechanisms on the RC use an articulated control feature?

None, I wanted to control other areas of the craft also, but this would have added too much weight ie. stepper motors, additional power, and costs to this junkyard project.

After I see that this will work, then I can start on version 2 with all the goodies

I think that you must have an active force control in the horizontal plane since friction, in this plane, will be extremely low and any force in this plane will generate a displacement of the platform. If the surface you run over is not horizontal the platform will follow the slope and "fall". Using 4 pads there is a further problem you may have to solve: on a not plane surface one of the pads will be either higher or lower with respect to 3 others. Such differences affect the gap and thus the pressure under the pad and the position of the platform. Using 3 pads the platform will take the position wich leads to an equal force distribution (all gaps equal) and will change its angle with respect to vertical generating forces in the horizontal plane. Due to small angles the forces are also low and may be only an air jet can compensate them. But for it a double angular sensor has to be used for stabilisation.

If the surface you run over is not horizontal the platform will follow the slope and "fall"

As in all hovercrafts a constant steering/counter steering is required to maintain position control over the craft. Mounting a servo to each pad at the A-Arm connection should give me 90° left/right control.

Using 4 pads there is a further problem you may have to solve: on a not plane surface one of the pads will be either higher or lower with respect to 3 others. Such differences affect the gap and thus the pressure under the pad and the position of the platform.

That is why each pad will be attached to the main frame using an A-Arm assembly to permit individual suspension allowing for each pad to be in a higher or lower plane as the others.

Amilo