Using Coanda effect for recumbent tailbox

I want to improve aerodynamic performance of a recumbent tailbox by smoothing the airflow around rider and reattaching it to tailbox by an enshrouding duct using Coandä effect - to forcibly channel the air against the tailbox to “make it stick” and provide pressure recovery for reduced drag without the need for a full fairing (which is hardly practicable).
Here is a VERY rough model:

My mockup model (admittedly, a mess) is getting rejected by the mesher, so I’ve created a very simple proof of concept model using simple geometric shapes - because I have NASA data with Cd values.

Here is my link to the project:

Unfortunately, the air seem to refuse to “stick” to tailbox, though I did get some improvement in Cd.
Does the system model Coandä effect at all, or I should make some adjustements? Making the system much larger/more complex would defeat its entire purpose and I’ll be better off with a full fairing perhaps.

Super nice project @OTsyganov!!

@Retsam, your project first came into my mind when thinking about the Coanda effect. Any impetus from your side here?



@OTsyganov: Interesting research. Coanda effect or not, it would not help you to reduce the drag. On your setup, shroud will add to drag in any case. You will soon find, that by adding fillet to the cone edge will be the main factor to reduce the drag. Even very modest fillet has big effect there.

When visualizing simulation results, use Particle Trace to see the flow. At those speeds it is mainly laminar.

Your mockup model is complex and it will take time to extract only essential parts and simulate flow with substitution forms for cyclist and cycle itself. Enjoy!



I see… why, though? Is it because the air is not “small scale turbulent” enough to stick to convergent tail section? Is there anything to be done about it?

Unfortunately, there is a limit how much you can “fillet” a human body :slight_smile: Of course, current “state of the art” dictate a tailbox that is a bit larger that the body and use that space to smooth the transition from body to tailbox, but it increases frontal area, I thought ducting will do the same…

By the way, is there an easy way to see surface pressure distribution on a model?

Please use Iso-surface or Iso-volume linked to pressure in order to visualize the pressure distribution around your shape.

My suggestion to ‘fillet’ was related to your cone shape from your project, not to human body. :wink: Turbulence develops with flow speed. Laminar flow is omnipresent and knowing which shapes allow for longer laminar flow is useful. Humans try to eliminate turbulence in their devices, but dragonflies know better! They use turbulence to master their magic fly in all directions. This is a small weekend digression… :man_student: