PDA

View Full Version : Fan noise - an indicator of efficiency?



jar2 - n/a
21-Nov-06, 07:44 PM
I was messing around on my Prospector hovercraft a few days ago and noticed an effect that has me a bit puzzled.



The Prospector has twin 24" 4z/9 lift fans driving into the main air cushion. I've always noticed that the fan noise seems to increase dramatically once the engine reaches about 2800rpm or so (fan speed is around 2200rpm). The fan noise signature changes from a fairly quiet whooshing to the typical fan howl you hear on a lot of integrated craft.



This particular craft is on full cushion at 2100rpm on grass. As an experiment, I set the throttle so that the fan "howl" had just turned on - I then opened a skirt vent to dump some air from the cushion. At this point the fan turned quiet again! I could turn the fan howl 'noise' on and off just by adjusting the size of the vent. There was no noticeable change in engine rpm during this time.



It's obvious that by opening a vent I'm increasing the air volume passing through the fan and cushion - however, the pressure and power consumed stays the same.



The question is - what's going on here? It would appear that the noise starts at the point at which the fan can no longer push more air into the cushion. The fan operating point is obviously moving but the pressure remains constant (the craft is still on full hover). The next obvious question is what is generating the noise? Is it turbulance around the fan blades caused by air moving back out of the cushion? Or is it something else?



Once there is an answer to this the next logical step would be to find some way to make sure the fan is operated below the point at which is starts to become noisy. This point would vary depending upon the cushion losses.



Comments would be welcome!

team black - n/a
21-Nov-06, 07:49 PM
If you watch a propellor in water you can see an effect called cavitation, where the prop spins without really biting at the water. It sounds like you may be describing a similar effect.



This makes sense when you think of the siren principle, with blades chopping the air to make a sound. I miscalulated the fan speed on a craft many years ago and ended up with something that sounded for all the world like an A10 tankbuster

jar2 - n/a
21-Nov-06, 08:17 PM
I though that cavitation was the formation of vapour or gas (or a vacuum) bubbles in a liquid? It shouldn't be possible in air (at least not at the kinds of pressures around fan blades)?



The fan is still shifting a fair amount of air - it just turns noisy. Maybe it is only one section of the fan blade that is making the noise - with the remainder still moving air?

Jeremy - n/a
21-Nov-06, 09:25 PM
John,



It sounds like the fan is stalling due to the flow conditions when run above that critical rpm, almost certainly caused by the fan dynamic back pressure (for want of a better term).



My experience is all with propellers, but they exhibit precisely the same symptoms when run at, or close to, blade stall. The most common occurrence of this on aircraft is when the inflow velocity is too low for the set blade pitch and rpm. As soon as the inflow velocity rises, as the forward speed increases in an aircraft (equivalent, perhaps, to the increased mass flow rate you get when opening a skirt vent) the blades unstall and run quietly.



This may or may not be the cause of your symptoms, but the easy way to prove it is to just reduce the blade pitch. If this shifts the noise-making rpm upwards, then you know for sure that it's blade stall causing the problem.



Jeremy

Ian Brooks - n/a
21-Nov-06, 09:38 PM
Hi



I don't know whats going on - but here's an experiment you could try:



As the fan gets faster, with no extra air flow, the angle of attack of the airflow to the blade changes. I wonder if the airflow is separating from the blade (stalling)? If this was so, reducing the pitch should move the 'howl' point upwards.



I here's an interesting statement - seems to fit your symptoms?



"stalled fan continues to deliver air, but at an increased static pressure and a decreased volumetric flow rate, and also at the cost of an increase in noise."



It's from this site:



http://www.electronics-cooling.com/Resources/EC_Articles/MAY (http://www.electronics-cooling.com/Resources/EC_Articles/MAY96/may96_01.htm) 96/may96_01.htm



Isn't internet remarkable??



This is indeed an interesting thing... if this noise is due to stalling... ("...it sounds "like a typical integrated craft...") then maybe we have a possible explanation why propellers seem quieter than fans - they are carefully designed so that they do not stall.



Food for thought here - we really do need to understand the velocity vectors, I can feel another program coming on... (but not until the Surveyor is finished!!!)



Cheers

Ian

jar2 - n/a
22-Nov-06, 09:13 AM
The blade stall theory sounds good to me. I am not keen on changing the blade pitch on two 9 blade fans due to the work involved! However, I did carry out another test that probably has the same affect - I increased the engine rpm above the 'noisy' rpm with the skirt vent opened (to stop the noise) and the fan noise started again at about 500rpm above the previous level.



The next question has to be what is going on on an integrated craft where there are two different exit areas behind the fan. A low pressure thrust zone and a higher pressure lift area below the splitter. It may be that the blade is stalling as it passes over the splitter area due to the reduced air flow. If this is the case then the turbulence around the start and end of the splitter must be pretty bad. There may also be greater blade deflection over the splitter - could this continuous cyclic stress be an explanation for some of the blade failures on integrated craft?



Too many theories and not enough facts!

Keith Oakley - n/a
23-Nov-06, 03:21 PM
John

Sounds like fan stall to me - I noticed a similar effect last year when analysing the overall dba and frequency pattern produced at different blade angles at the same rpm on my F25. If you can record say a 10 sec wav file with the vent opening and closing for a few secs each time then email me the file I'll try to frequency analyse it and see what the differences are.

I agree that a lot of specific sounds in the hovercraft medley are generated by various forms of turbulence and thus sounds are a pointer to various inefficiencies. - the challenge is to crack that code!

Keith Oakley

jon_curtis - n/a
3-Dec-06, 08:29 PM
i have been reading an artical about thin aerofoil stall,



to sum it up. (it might all be rollocks lol)



as the aerofoil reaches stall, laminar seperation starts on the curved portion behind the leading edge, in seperation this flow becomes turbulent and if it has the energy turbulent reattachment follows.

trapping a small seperation bubble on the aerofoil surface, increase in speed or angle of attack will cause this bubble to move aft along the aerofoil untill it unattaches and the aerofoil is in full stall.



what is interesting is that, if turbulence is introduced on the trailing egde of the aerofoil, it can introduce enough energy to reattach the bubble, this will have the effect of increasing the speed or angle of attack at which the aerofoil will stall.



could you do us a favour john and stick a couple of layers of sparky tape 2/3rds of the chord on the fan blades along the lenght, to see if it will raise the rpm at which this happens?

jar2 - n/a
4-Dec-06, 06:55 PM
Keith, Jon



Once the weather settles down I bit I'll try the tests you suggest.



I don't think I have a specific problem with my setup (any different to any other fan setup that is!) - it's just, as Keith says, trying to understand what is happening in the hope that I can learn something!



These blades are set to 25degree pitch - from memory, this is pretty near the minimum (20-22degrees?) before the tips end up with a negative pitch.



A simple way to keep the fans operating in the 'low-noise' mode would be to have an automatic cushion dump valve (pointing backwards to give a couple of pounds extra thrust?). Simple enough to implement apart from the valve control - how could you detect the onset of blade stall (assuming that's what it is)?

jon_curtis - n/a
4-Dec-06, 07:03 PM
if you run smoke into the fan you should be able to see it, well it works in a wind tunnel.



you should see the turbulent flow forming!