Low speed (<25mph) Aerodynamics

[Sven7]

My friend and I got into a discussion at the EcoMarathon about low speed aerodynamics. We know that the airflow changes at lower speeds, but we don't really know how.

Most of the prototype cars there are more or less teardrop shaped tadpole trikes, like what you would design for a highway-speed car. He thinks that you could basically build a delta trike with a conical body that was chopped off at the back. Is this the ideal design? What is the ideal aero shape for low speed vehicles?

There doesn't seem to be much info out there about this stuff. Hopefully someone will have an idea!

Thanks in advance!

[aerohead]

Originally Posted by Sven7

My friend and I got into a discussion at the EcoMarathon about low speed aerodynamics. We know that the airflow changes at lower speeds, but we don't really know how.

Most of the prototype cars there are more or less teardrop shaped tadpole trikes, like what you would design for a highway-speed car. He thinks that you could basically build a delta trike with a conical body that was chopped off at the back. Is this the ideal design? What is the ideal aero shape for low speed vehicles?

There doesn't seem to be much info out there about this stuff. Hopefully someone will have an idea!

Thanks in advance!

For the lowest drag,you need a supercritical Reynolds number.
If the body is too short to achieve this,you'd have to introduce artificial roughness to trigger the transition to a turbulent boundary layer.
Otherwise,the flow would separate before it ever got to the point of maximum cross-section.
Trip wires,sand,VGs,and dimples could be used for the trigger.After that,then it's either a wing-based body,or pumpkin seed for the lowest drag.Whatever the driver could tolerate.
Cones have full-frontal area-wakes.There's no pressure recovery possible with them.They have double the drag of a reversed-cone with convex hemispherical nose.This data dates to as early as 1898 with Alexander Gustav Eiffel's research at his famous tower in Paris.They violate the entire premise of streamlining.You want full boat-tailing if you can do it.Reduce or eliminate the wake.
Here you see the effect of low Reynolds number and enormous wake

[Sven7]

Ok, I'm trying to wrap my artist head around this. If I am reading this correctly:

-A wing or pumpkin seed shape is more efficient at low speeds than a conventional teardrop

-It is advantageous to add flow trips at the nose to reduce wake size even though the wake is now turbulent

-"supercritical" in this context means that the wake is smaller than it would be on a "normal" (figure A) shape

-There is no merit to the backward-cone theory

Is that all correct?

PS-
Corollary: does this indicate that Paul Jaray meant for his Ley T6 "pumpkin seed" to go very low speeds?

[aerohead]

Originally Posted by Sven7

Ok, I'm trying to wrap my artist head around this. If I am reading this correctly:

-A wing or pumpkin seed shape is more efficient at low speeds than a conventional teardrop

-It is advantageous to add flow trips at the nose to reduce wake size even though the wake is now turbulent

-"supercritical" in this context means that the wake is smaller than it would be on a "normal" (figure A) shape

-There is no merit to the backward-cone theory

Is that all correct?

PS-
Corollary: does this indicate that Paul Jaray meant for his Ley T6 "pumpkin seed" to go very low speeds?

*anything with a blunt,rounded nose and fully-tear-dropped tail,whether wing,or half-body of revolution will be good.
*to be the lowest drag the wing will try and be laminar,with its thickest point moved as far rear as possible to delay the transition to a turbulent boundary layer (to minimize skin friction).If there is any turbulence in the ambient air,or if there is any crosswind,the wing will transition directly to a turbulent boundary layer.But this is okay,as it will be impossible to have separation.
*to be the lowest drag half-body, it will need a supercritical Reynolds number or else the flow will separate before it ever reaches the max camber point along the body and never have reattachment.
------------------------------------------------------------------------
So the Reynolds number is critical if the bodies are very short and going really slow.

Reynolds number = Length (L) X Velocity (V) / kinematic viscosity (v)

or, Rn = L X V / v
You can use 0.00015 ft-sq/second for v
1/v = 6380
so Rn = L X V X 6380 as a easy working number.
--------------------------------------------------------------------------
With turbulence in the ambient air,the TBL transition occurs around Rn= 75,000. This would be your minimum target based upon your length and velocity on the race course.
-------------------------------------------------------------------------
If your Reynolds number comes in low you need to roughen the nose to force the TBL transition,then you can get attached flow to the rear.
--------------------------------------------------------------------------
Supercritical refers to the Reynolds number which allows the transition to TBL.(around 75,000)
--------------------------------------------------------------------------
The cone,pointed end into the wind would have 54% higher drag as the cone's point headed backwards IF it has a convex-hemispherical nose.(think of an ice cream cone of 20-degrees angle:Cd 0.1698 vs Cd 0.0925.
Pointed end first,with the big rear end won't allow any deceleration of the flow before separation,which means low pressure in the wake,which increases the pressure drag,which violates the fundamentals of streamlining.
--------------------------------------------------------------------------
Jaray was designing for passenger cars at normal driving speeds.His pumpkin seed would be low drag at any speed as long as the Reynolds number was supercritical.This is why we have to be careful with scale models in a wind tunnel.If small models are not run at sufficiently high airspeeds,then they may never reach TBL,and the Cd may bounce all over the place.

[Grant-53]

It may seem odd but based on the bowling balls I might try sanding the leading edge of my spokes. Any suggestions on the grit number that would be best?

[aerohead]

Originally Posted by Grant-53

It may seem odd but based on the bowling balls I might try sanding the leading edge of my spokes. Any suggestions on the grit number that would be best?

I'll look at Schlicting's book.If it's anywhere,it will be there.

[MobilOne]

What I deduced from the above is that it is nonproductive to wash and wax your vehicle. LOL

[Sven7]

Ok, so in short, the ideal aero shapes are the same as we're used to, but the air just behaves differently around them. Thanks

[aerohead]

Originally Posted by Grant-53

It may seem odd but based on the bowling balls I might try sanding the leading edge of my spokes. Any suggestions on the grit number that would be best?

I found some info in my fluids text.It's quirky!
Spokes very in drag coefficient,from Cd 6.0,to Cd 0.30,depending on Reynolds number.
The drag minimum occurs when the Rn is at least 1,500,000.
--------------------------------------------------------------------------
I measured the spokes in my mountain bike,3/32nds-inch.
If you plug this into the standard Reynolds number formula and solve for critical velocity,it's beyond the speed of light,given a smooth spoke.
--------------------------------------------------------------------------
Looking at artificial roughness,and critical roughness we're given:
Critical Roughness (Epsilon) = 15 X (kinematic viscosity/ [Square root of the Shear Stress / air density (rho)].
the equation is simplified to :
Epsilon = 26 X [( kinematic viscosity/ Velocity) X Reynolds number raised to the 1/4th power ].
-------------------------------------------------------------------------
If we plug in Rn= 1,500,000 and Velocity =, say 15-mph (22 ft/sec) and solve for Epsilon,we get 0.0005-inch as the critical roughness.****************
I'd sure like some independent verification on this!
-------------------------------------------------------------------------
I went into my Technical Aerodynamics text and pulled out the following:
*A 0.375" round spoke has 0.0054 pounds drag per foot at 15-mph
*A 4:1 elliptical spoke of 0.375" has 0.00414 pounds drag per foot at 15-mph
*A 4:1 streamline spoke of 0.375" has 0.00108 pounds drag/ft at 15-mph
The streamlined spoke has 80% less drag than the round spoke.
Seems like the spokes should be smooth and streamlined.And minimum in number.
--------------------------------------------------------------------------
In a calm,the spokes would be in train behind the tire/rim,embedded within the rims turbulence.
In a crosswind,the air would be hitting the spokes at a spectra of angles,complicating streamlining.
--------------------------------------------------------------------------
I would be tempted to fully-enclose the wheels as with a solar racer,and create shutters which were spring loaded to open instantaneously in dangerous crosswinds and forget about trying to streamline the wheel assembly itself altogether.

[Grant-53]

There has been a number of papers generated on the bicycle wheel and a variety of spoke designs. Weight is always an issue for human powered vehicles. I have had some good results using a smaller 20" front wheel with the right head tube angle. I will get some measurements on spoke diameter and length. Thanks