Ultra low drag vehicles
 

I don't think I have linked to this here before.

If you're after ultra low drag and you are starting from scratch (velomobile, unique car, solar car, gravity car, unique RV) here are some things to pay attention to.

https://www.youtube.com/watch?v=iSIpvOr3w3I

In addition to the solar racers look at the velomobiles of the IHPVA on recumbents.com and the motorcycles at craigvetter.com. High performance sailplanes are very efficient as well.

Questions:

• Is the order important, i. e. attached flow and wetted areas more significant than frontal area and interference drag?
• With an open wheel vehicle (my favorite use case) is the interference drag between the wheel/tire and the body or is it between the axle/suspension/steering linkage and those?

How well would a Dymaxion [Omni-directional Transport] with a boxed cavity truncation fare against these guidelines?

The reference I am quoting puts them in the order of importance shown in the video. In the context of the video - neither. It refers to where different body parts join eg wheel spats to main body.

You can guess that as well as I can!

I'm in the process of re-designing my velomobile. This new design will be the third iteration.

Are there any free tools that you are aware of that will aid in the design of the shell, which will give actual charts or numbers to work with and not just a graphical representation? I'm looking for something that tells me what the air is doing as it passes over the shape, whether the flow is laminar or turbulent, whether the boundary layer is attached, what the pressure is at various points in the shape, overall drag coefficient value, coefficient of downforce/life values, ect.

This video was a good basic primer, but the 6 actual steps listed to minimize drag, little was explained regarding exactly HOW it is done.

I don't have thousands of dollars for software or tens of thousands of dollars for wind tunnel testing, nor do I want to have to make multiple iterations of the same thing as that takes hundreds of hours of work each time just to get a chance to tuft test it.

The short answer is: no there isn't. If there were, car companies wouldn't spend millions on computers and aero software, and hundreds of millions on wind tunnels.

This is one occasion when copying one of the generic low-drag shapes will probably get you 90 per cent there.

Wouldn't spats apply to a mono-hull? I'm thinking of three or more aeroforms flying in formation, connected by gnarly beams and struts.

This is why I liked Oliver Kuttner's in-wheel suspension and steering. It would allow for a properly filleted intersection.

https://ecomodder.com/forum/member-f...11-5-38-12.png

The Toecutter — Blender has an Export To Paper Model function that will give you a cutting pattern for an arbitrary shape, but I haven't seen convincing CFD yet.

As we've previously discussed, I am loath to just guess stuff. And, since I have no tech literature on this subject, and haven't personally tested it, I have little idea.

free tools
 

The International Human Powered Vehicle Association ( IHPVA ) would probably be your best bet. They represent the international community of university engineering departments, which come together south of Battle Mountain, Nevada for land speed record attempts, on a section of state highway, intermittently closed to through traffic by state police for the event.
They run fully-enclosed streamliner bicycles, and three-wheel velomobiles.
Battle Mountain's convention center is a hive of activity during qualifications and competition.
Lot's of Cd 0.11 bikes and 85-mph on a good day! It's a real hoot to attend. It used to be $ 35/ year ( US ) to join, and receive newsletters.
Check out their website. Or attend, and take in Bonneville, during the same week.:)
Last time I attended, one of the amateur members did 57-mph in his Coroplast-clad 3-wheeler.

At recumbents.com there is the shell design program that can be used to create a shell form. The file then can be used in CFD programs such as AutoDesk. The question as with all measuring devices is how good is good enough?

The Ford Probe was an example of a low drag prototype that inspired a production vehicle.

Is?
 

1) Your 1st question isn't specific enough. We need some caveats, conditions of the vehicle specification in order to work with it.
2) The 1976 CNR project provides insight into the DYMAXION Car question.
* The Morelli body of the 'Banana' car was Cd 0.161.
* When wheels were added, the Cd jumped to Cd 0.35.
* When the wheels were 'integrated' ( Hucho's terminology ) better into the bodywork, the drag fell to Cd 0.201.
* Hucho has also written on the remarkable drag contribution of the open wheels of the Rumpler Limousine, measured at Volkswagen.
* The 1987 GM Sunraycer was reported to have a difference of Cd 0.089, and Cd 0.12 when the full wheel-fairing package was removed for the World Solar Challenge.
-------------------------------------------------------------------------------------
As to the DYMAXION Car box cavity. Again, we'd need to know how much truncation would be performed, and then the type, and length of the box-cavity.
* Hucho depicts one type of architecture.
* United States Patent # 4,682,808, Filed July 24, 1985, by Alan J. Bilanin depicts a proper cavity. His twenty-one references date to as early as 1933.

monoposto
 

I believe that we explored this situation over a decade ago. Around 1958, Mercedes-Benz published that, in spite of its significantly greater frontal area, and weight, that it's fully-enclosed Grand Prix race car had significantly less overall drag than the same car, as an open-wheeled race car.
And yes, it has to do with the interference drag of a 'mutilated' ( Frederick Lanchester's terminology ) form.

how good
 

The 2015, ECORUNNER - V, by Delft Technical University, may have reported the lowest extant drag for a land vehicle. Cd 0.0512.
This would have to be considered within the context of a 3.8-sq-ft frontal area vehicle, averaging 15-mph ( burn - and - coast ), on a closed-course, perhaps completely shielded from any wind, or terrestrial boundary layer.
It is a 'three-wheeler', so, it might qualify for a 'fairing' on a velomobile.
The 2006 PAC car-II is also a three-wheeler, and formally held the low-drag record at, Cd 0.07. Perhaps at a higher Reynolds number than the ECORUNNER-V.
Only you can judge how far you want to push the CdA. And under what conditions.

Probe
 

Yes, the PROBE-III. Cd 0.22. Which influenced the 4-Door, Ford Scorpio/ Merkur XR4Ti, Cd 0.34 w/o bi-wing spoiler, Cd 0.32 with.
The shorter, two-door sedans were Cd 0.33 with spoilers.
The Ford 'PROBE' production car, constructed at the joint-venture Ford / Mazda facility, took the 'PROBE' name, but none of the concept car's architecture.:(

EXACTLY how would you shield from the terrestrial boundary layer? If it flies, it isn't a car, ditto for a vacuum.

What is meant by that, I think is that there is no wind across the road, the terrestrial boundary layer would be there if there is wind. It was a complicated way of saying, done on a windless day.

shield
 

All one could do is not drive when there's wind, or solar convection currents present. That would limit the UK to 7-days/ year according to MIRA.
Aircraft are designed for 'flight' conditions, above Earth's TBL.
Aircraft do experience winds aloft, convective columns, Katabatic winds, the Jet Stream, clear air turbulence, and such, so they do have 'issues.'
Just not 'automotive' issues.:p

I did that already, and it's definitely not 90% there, maybe 30% there, although I think that has a lot to do with the fact that I still have outboard open wheels. Due to the clearance issues posed with the steering geometry and suspension, wheel wells were difficult to implement. I wanted to avoid having large wheel well gaps and needed to retain access to the brake calipers and axle for maintenance/repairs(often on a sidewalk 30+ miles from home), so the first two shells had open/outboard wheels. And that is really bad for drag.

I made some farings for the front wheels out of aluminum, but they actually added drag. I didn't have them shaped right and their frontal area penalty likely outweighed any drag coefficient reduction they provided for the wheels, plus the backs of the farings seemed to act as parachutes. I over-sized the clearances for them because I did not want them to give me a flat tire, and even then there was a bit of rub that had to be corrected!

I'm working on a set of more narrow profile farings that may actually show an aerodynamic benefit. We'll see.

I think I solved how to shape the wheel wells for an inboard wheel design with an inverted half-toroid shape, which should hopefully allow operation without any tire scrubbing during cornering or over rough roads, for the next design iteration. If successful, I expect greatly reduced overall drag.

how would you
 

What immediately comes to mind would be a race course, buffered from ambient wind, embedded within the stadium, high crash-wall, environment. Like The Indianapolis Motor Speedway, or Phoenix International Speedway where I observed the Solar 500.
A laminar aircraft is designed for 'flight' conditions, up and away from terrestrial effects, such as Earth's boundary layer.
Hucho depicted an Urban boundary layer extending to around 525-meters AGL.
According to MIRA, there are only, on average, 7-days a year in England, free of terrestrial turbulence.

Ok you are not shielding from the boundary layer, you are minimizing the adverse effects. Ideally you want it driven/measured in an airconditioned building?

shielding
 

A thought experiment to illustrate the race course context would be:
* The raceway is in the 'lee' of a hill ( Phoenix Int'l Raceway ).
* The hill behaves like a the cabin of a pickup truck.
* The race course is embedded within the stagnation bubble 'behind' the cabin.
* There's a Prandtl line of discontinuity created in which the race course resides, as long as the wind doesn't shift.
* Ambient conditions observed on the race course are 'dead air.'
* A small, sub-critical Reynolds number race car could get away with a 'laminar' body under this condition, as long as its velocity was held below the sub-critical Rn transition velocity. We're talking frontal areas of 3.8-sq-ft.
* No 'full-size' car, at typical travel velocities could ever get away with that.:p

I see this as a moving collection of variable conditions which may or may not average out. Then there is the "laminar" body.......what part of this body is laminar and how much of the turbulent aft is going to affect data since I have yet to see any shape 100% laminar. I can see a very low drag shape getting consistent data but that is the limit, I believe. I also envision the reynolds number wandering around since race tracks have straight-a-ways to go fast on and tight turns to slow on which will affect the turbulence unless you pick a slow enough speed as to not accelerate or brake.

'laminar'
 

Good points!
* 'laminar' is really an oxymoron.
* There's no such thing.
* What passes for 'laminar' is, a surface of laminar boundary layer, up until the flow experiences the first minimum pressure, at maximum body cross-section, then from there to the trailing edges it's all an adverse pressure gradient where it's impossible for a LBL to exist, and the transition to TBL happens immediately downstream of this maximum thickness.
* The context of 'laminar' bodies as seen at SHELL Eco Marathon, SAE, and World Solar Challenge events lies within the frontal area ( from which Reynolds number is directly associated ) and 'race' velocities, which are typically quite low. And 'calm' conditions.
* And yes, the Rn will vary with velocity. 15-mph ( burn-and- coast ) average speeds are not uncommon, and a 22-mph upper limit according to rulebook.

All true except 'laminar' can't be an oxymoron.

can't
 

How about 'misnomer' ?
:)

No, this is not true! Reynolds number is directly proportional to speed and characteristic length, not area.

Rn
 

Characteristic 'length' for automobiles is derived from frontal area.

That's seriously going from the sublime to the ridiculous. It seems you're now just making stuff up - no expert, reference or paper supports this nonsense.

Fineness ratio?

Has nothing to do with Reynolds Numbers.

Rn
 

I never said I like it. I'll dig through my rat's nest and see if I can trace it down.
I've made transcription errors before.
Thanks for flagging it.:)

I suppose you could make an argument that frontal area and length are correlated, but even that is only generally true. Take a car like the Smart Fortwo (A = 21.0 square feet) and the original Honda Insight (A = 19.8 square feet) and it falls apart; the Smart is only 106" long while the Honda is 155".

In vehicle aerodynamics we're restricting Reynolds number calculations to incompressible flow; once you introduce things like compressibility and heat flux there are many, many other ways to calculate it. Perhaps you're thinking of one of those?

It just seems very strange to me that you endlessly quote Hucho 2nd edition in support of your theories, but get wrong something as basic as how Reynolds Numbers are calculated.

Hucho 2nd ed page 49:

[The Reynolds number] is a function of the speed of the vehicle, the kinematic viscosity of the fluid and a characteristic length of the vehicle eg its total length as in Figure 2.2.

(And Fig 2.2 shows vehicle length exactly as we'd expect - front of vehicle to back of vehicle.)

No mention of area, and certainly no mention of working out length from area.

Incidentally, for people testing and modifying the aero of their car, Reynolds Numbers can basically be ignored. They only become relevant when different size models are being used. The only time I've ever had to pay attention to Reynolds Numbers is when calculating wing downforce / lift from wind tunnel aerofoil data.

a 'laminar' take on Reynolds number
 

1) Cd is based upon frontal area
2) Drag minimum occurs @ CdA minimum
3) 'Laminar' profile Cds are Rn-dependent
4) Rn is 'size', 'velocity', and 'kinematic viscosity'- specific
5) The metric for Rn-related ' size' is length
6) The 'length' of any given 'laminar' profile is a derivative of Af ( scaling-factor)
7) Competition 'laminar' bodies are 'shrink-wrapped' over, and constructed to fit around the body of a single individual team member chosen for 'pilot'
8) The smallest Af = the shortest length
9) The shortest length = smallest RN
10) The smaller the Rn, the longer the delay in reaching transition to super-critical Rn
11) Super-critical Rn brings a concomitant transition to TBL
12) TBL renders a 'laminar' profile impotent
--------------------------------------------------------------------------------------
A curiosity in research reporting is, that for the Rocketail Wing.
The investigators at the University of Glasgow derived their Rn of 5.3 X 10-to the 5th, upon the 18-wheeler's height, rather than length. First I can recall, ever.

1) cD is based on frontal area only for convenience' sake. We could relate cD to plan area, as in aeronautics, or wetted area, or any other reference area we choose. However, those are all much harder to measure or calculate on complex three-dimensional bodies such as cars.
6) "Scaling factor" in aeronautics and aerodynamics refers to the necessity of obtaining the same Re in the flow around a model as the full-size object, as in this paper written by some researchers working down the street from me. It appears you are conflating "scaling factor" with "fineness ratio," as freebeard suggested. But, since production and concept cars have widely varying fineness ratios this isn't an accurate method of predicting Re for cars based on height or frontal area, as I pointed out in a post above. (It might work for 18-wheelers, which have standardized trailer lengths, typically 53').

convenience
 

1) Frontal area-based drag coefficient has been the industry standard basically all along.
2) I don't mention 'aeronautics' just as I never mentioned 'aerodynamics' with respect to static 'lift'. Please read for comprehension and pay close attention to actual language!
3) Any 'profile' will be profoundly affected by fineness-ratio. Any given frontal area demands a matching length to preserve the fineness ratio.
4) For road vehicle aerodynamics, scaling-factor has to do with dynamic similarity, verisimilitude, precisely related to Reynolds number. It's an absolute necessity with respect to scale-model wind tunnel testing.
5) I don't 'conflate.' The science speaks for itself.
6) A 'life-size' automobile, above 20-mph, will be at super-critical Reynolds number, and constant drag coefficient up to transonic velocity.
7) We don't need to concern ourselves with RN effects of a 'real' car.
8) The context of modern 'competition laminar' bodies is exactly related to frontal area, as it undergirds every other dimension of the body architecture.
One cannot understand the 'context' of a 'laminar' body without this understanding.
That was the condition of the discussion.

I'm not even going to bother to respond to the rest of this because I'm not nearly as patient as some of the other posters here, but re: 7) One thing Hucho did write to me was an explicit warning to be careful in testing devices on my car since front and rear spoilers can be sensitive to Re effects.

Perfect example of Aerohead's approach.

1. Aerohead made a mistake in a post (associating the calculation of Reynold's Numbers with frontal area, not length).

2. The mistake was pointed out by Vman455, with a simple and clear correction.

3. Aerohead immediately doubles down, reiterating the mistake.

4. A person mislead by the Aerohead's posts appears, trying to make sense of what Aerohead is saying (rather than simply seeing it as wrong) and bringing in another issue of no relevance.

4. I also point out that what Aerohead is saying is wrong (but I am less polite than Vman455, because I can see yet another example of Aerohead breeding confusion on this subgroup - see above).

5. Aerohead agrees to consult his references to find out how he arrived at his error (an error he hasn't yet admitted, despite it being as clear as saying that force = mass * velocity)

6. Aerohead comes back with his 'wall of noise approach', an approach that mixes irrelevancies, illogical jumps and correct definitions in another amazing mish mash that attempts to justify what is a simple error on his part.

7. Vman455 tries to correct the new errors that Aerohead has introduced in his response.

8. Aerohead replies with another wall of noise, but this times moves the goal posts right away from normal road cars into laminar flow bodies, further sowing confusion.

9. Vman455 understandably starts to give up in his corrections.

Yet another example of where something as simple as this equation:

Reynolds Number = (density * speed * length) / viscosity

...has under Aerohead's guidance become an incomprehensible mix of rubbish, errors, confusion and irrelevancies.

Are we talking about Permalink #30. If so, then get stuffed.

I was pushing back against The Template long before you were around. The difference is I attack the concept, not the person. You mis-read a request for clarification.

Did I push back on that? I think there is some correlation.

So what is fineness ratio again?