Boattailing the Prius (nope, I'm not dead)

[MobilOne]

I think that they are both beautiful. They remind me of Alfa's from olden times.

[ennored]

Originally Posted by freebeard

Thee Template is more or less a guideline, as it assumes a half-circular cross-section. Taper in plan and tumblehome affect the result (see the GM EV-1). You can't even trust tufts completely, but Toyota seem to know what they're doing.

Something else I see missing around here...

AST (I or II) ≠ attached flow ≠ ideal

That is, just because there is attached flow, doesn't make a shape ideal. And you don't have to follow the template for attached flow.

I did some playing a bit ago with the template. As an engineer (being nothing more than a cup of coffee and a spreadsheet) I fired up Excel. Plotted and fussed with the template. I arrived at this:



At this point, the math/numbers are irrelevant. It let me draw the template shape (in 2D obviously).

We know that a sphere in a wind tunnel has attached flow for a few degrees past tangent. It's often expressed as 3 or 4 seconds on the clock face. Knowing that, and recalling the oft mentioned 22° (as happens to be found on "the template" too.) I drew this:



What do you know, 22° lands at 3.67 seconds. Hmmm....

Is this second shape a "minimum" shape to achieve attached flow? Data I've seen suggest it is. Or at least very close. Would this shape be ideal? No, we already know that. But, it ends up pretty close to the template, and perhaps explains why many "less than ideal" tails do so well. Here they are together:



Not all that different.

SO.

Less than ideal tail for a Prius, that follows the Toyota lines? It'll be less than ideal, but very good chance it'll exhibit attached flow.

[freebeard]

aerohead can speak to his experience renting time in the Darko windtunnel.

I need to find a decent cheap matchbox Prius and mod that first >.< lol

Matchbox models are 1/64th scale. Manufacturers use 1/4th or 1/5th. A reasonable compromise would be 1/25th to 1/16th. See http://ecomodder.com/forum/showthrea...ies-31047.html, where models are suspended upside down in a water tunnel.

[aerohead]

Originally Posted by ennored

Something else I see missing around here...

AST (I or II) ≠ attached flow ≠ ideal

That is, just because there is attached flow, doesn't make a shape ideal. And you don't have to follow the template for attached flow.

I did some playing a bit ago with the template. As an engineer (being nothing more than a cup of coffee and a spreadsheet) I fired up Excel. Plotted and fussed with the template. I arrived at this:



At this point, the math/numbers are irrelevant. It let me draw the template shape (in 2D obviously).

We know that a sphere in a wind tunnel has attached flow for a few degrees past tangent. It's often expressed as 3 or 4 seconds on the clock face. Knowing that, and recalling the oft mentioned 22° (as happens to be found on "the template" too.) I drew this:



What do you know, 22° lands at 3.67 seconds. Hmmm....

Is this second shape a "minimum" shape to achieve attached flow? Data I've seen suggest it is. Or at least very close. Would this shape be ideal? No, we already know that. But, it ends up pretty close to the template, and perhaps explains why many "less than ideal" tails do so well. Here they are together:



Not all that different.

SO.

Less than ideal tail for a Prius, that follows the Toyota lines? It'll be less than ideal, but very good chance it'll exhibit attached flow.

You're barking up the right tree.We went over this a couple years ago with 'Rolex meets Big Ben'.
*The TBL golf ball established the separation-free contour out to 4-seconds

*I held the forebody constant,and then began increasing the radius of the aftbody until the entire span was the 'midnight-to-4-seconds after' contour.to create a default 'Template.'
--------------------------------------------------------------------------
*It's extremely important that the contour leading to the 22-degree tangent be no 'faster' than it is with the 'Template'.The boundary layer is already in a hostile environment and if you throw too much curvature at it,the rapid pressure rise will trigger separation,and they'll be no reattachment.
*Here's W.A.Mair's wind tunnel-tested aft-body.Notice the very gentle lead-in to the 22-degree slope.

*According to boundary layer theory we can't cheat on this,that's why the 'Template' is as 'slow' as it is.
*A 2.5:1 fineness ratio streamline body of revolution has the only contour which both minimizes separation,while also minimizing surface friction (wetted area).
*This 2.5:1 body is the basis of the 'Template.'
*But is all you have to go by is an analog clock or watch,it's not a bad contour.

[ChazInMT]

ennored, you're right about attached flow ≠ ideal. People seem to think as long as you have attached airflow, you've optimized the Cd. Not the case. You need to have an even balanced pressure return in the back of a car, this is the most critical aspect of trying to optimize a cars shape. If the shape follows the template while maintaining balanced pressure all around, it will have the lowest Cd possible. If the shape is faster/steeper/shorter than the template curve, it will suffer from creating too much of a low pressure area, and the Cd will rise. If it is slower/shallower/longer than the curve, it will start to suffer from more skin drag and on a more esoteric level, the air will not help to push the car along with the same force as it returns the prior state it was in before the shape rammed through it, again the Cd will rise.

So yer right, if one part of the car follows the template shape, and the other parts do not, it isn't ideal.

But it is a fantastic starting place. It is a tool that if used properly will get people closer to an optimized shape for a given form. It is a well thought out basic tool, not the be all-end all for aerodynamics.

If you have a better tool for us we'd love to see it and discuss it. Do you have something that is a more aerodynamically ideal shape for a bluff body moving in ground effect?

I been looking into this for 5 years now and think about it quite a bit, and to me, the templates are as good as it gets for a starting point based on all the research that's been presented by Aerohead over the years.

[ennored]

I guess I did oversimplify a bit when I wrote that the template isn't ideal. It is, but we rarely (never?) see it implemented perfectly. We usually have a base shape (a car or trailer) that doesn't match the template. The template is often applied in 2D and assumed to give the right shape (of a tail or other feature) and, as has been mentioned, that isn't enough. And neither is attached flow. That state it better?

[aerohead]

Originally Posted by ennored

I guess I did oversimplify a bit when I wrote that the template isn't ideal. It is, but we rarely (never?) see it implemented perfectly. We usually have a base shape (a car or trailer) that doesn't match the template. The template is often applied in 2D and assumed to give the right shape (of a tail or other feature) and, as has been mentioned, that isn't enough. And neither is attached flow. That state it better?

GM was worried about cars that looked alike since the 1920s,and there's institutional inertia biased against form following function.Zero consumer education as well.
Upcoming CAFE standards for 2025 are going to push carmakers hard for mpg technology.
We may see closer 'fits' as far as ideal forms.
In 1985,Wolf H. Hucho mentioned that cars of Cd 0.13 were not a question of technological feasibility,but rather how important low drag was in the decision making process for production.
It should be an interesting next ten years.

[freebeard]

It should be an interesting next ten years.

Someday they will all look like this:

[ChazInMT]

Originally Posted by aerohead

Upcoming CAFE standards for 2025 are going to push carmakers hard for mpg technology.
We may see closer 'fits' as far as ideal forms.
It should be an interesting next ten years.

Sooooooo

Here's the 2025 Camaro




And the 2025 Mustang



And the 2025 Camry

[whatmaycome14]

LOL!!! That's too funny.