0

I have an axial 140 mm fan which I need to connect to a duct and I need to design a converging duct as adaptor from circular to the shape and size of the duct I have.

From past experiences, an axial fan followed by a duct will result in too low flow, the speed at the exit of the fan has too much swirl to be efficiently propelled forward.

In the past for other purposes (from fan to nozzle) I designed and 3D printed a converging duct with simple straight vanes forming a cross centred on the fan itself. The flow was better but still quite poor.

How can I roughly estimate, based on rotational speed of the fan and on the air flow of the fan, which angle of attack I should use for the vanes?

Of course an accurate calculation would require flow simulations, I am asking about estimates which can be obtained without having to simulate the fluid.

FarO
  • 169
  • 10
  • This has two sub-problems: firstly finding out the direction of the outflow from the fan (as a function of radial co-ordinate), and secondly designing a vane that will turn that flow to get rid of its azimuthal component. Am I right in thinking that you already know how to do the second part, and you're only looking for help with the first part? –  Jan 16 '24 at 14:31
  • Build a test rig where you can alter the vane angle and test - likely to give better and faster results. – Solar Mike Jan 16 '24 at 17:47
  • @DanielHatton I think I can do the second part, also I saw how vanes are shaped in existing fans. It's the estimation of the swirl which I'm not sure about – FarO Jan 16 '24 at 20:38
  • @SolarMike well designing a rig with variable vanes doesn't seem that easy either – FarO Jan 16 '24 at 20:40
  • @SolarMike There won't be a vane angle, the vanes will need to be curved. –  Jan 16 '24 at 21:16
  • 1
    @FarO You could try and compute the angle of the output flow using the flow-directions analysis that's behind the Euler turbomachinery equation (see Douglas et al.'s Fluid Mechanics for details), but the Euler turbomachinery equation usually gets the energy input to the fluid wildly wrong, so I wouldn't have much confidence in the results from that. Best bet's probably to measure the output direction - maybe tie some thin paper streamers to the output side of the fan housing and see which way they point. –  Jan 16 '24 at 21:25
  • @DanielHatton I was more expecting a reply which would help me estimate the radial distribution of the flow, which would help me define the axial speed as function of radius, which together with the speed triangles would help me estimate the angle. Testing with paper streamers (to a steel wire stretched radially?) could also be an option indeed. If you summarise the equations just for the sake of "answering" and you add the tip you gave me together with my reply, in a week or so I would likely approve the answer – FarO Jan 17 '24 at 08:57

0 Answers0