Want low drag? Don't follow a template.

[JulianEdgar]

Originally Posted by aardvarcus

Julian,

Did I comment on your diagram use incorrectly? You are using the AST-I for all your overlays, are you not? In your videos and posts and such?

I am using the template that I see widely used here - the one in the Tools section of the site.

I really don't see that many users using that diagram (AST-I not II) on this site for those purposes.

Oh well, that's pretty well all I have seen being used! But it changes almost nothing: the use of any predetermined template to do any of these things is absurd:

- Show where there is separated and attached flow on existing cars
- Guide the shape of rear extensions
- Show how rear spoilers on sedans should be positioned and shaped
- Allow the assessment of the ‘aerodynamic purity’ of cars)

Using a diagram at all may be absurd, but in terms of relative absurdity I would consider using a diagram that exactly matches the XL1, Prius, Insight, etcetera to be less absurd than applying a diagram intended for a use on a vehicle that is in cross section a hemisphere. Once again, I don't have any hemispheres parked outside.

The template (whichever one you have now chosen) matches the font, side and underneath profiles of all these cars you've cited, does it? Er, no it doesn't...

If the notion of someone who does not have development resources mimicking the rear curvature as a function of height of some of the most respected low drag vehicles is absurd, then yes please count me among the absurd.

Well, it's just as absurd as blindly copying the full-load air/fuel ratio from another engine and saying it's good for your engine. Or spring rates, or....

I saw a really good quote on this the other day, it said "This is one occasion when copying one of the generic low-drag shapes will probably get you 90 per cent there."

Sure, I said that. Now why not give the context? The development was of a velomobile from scratch - not the modification of a car. In that context, yep, follow any of the at least five (!) low drag shapes that have been published.

The professionals may not use a diagram, but it is funny how so many of their "low drag" designs end up with almost exactly the same rear curvature as a function of height.

And so many of their low drag shapes do not. See a problem here!?

I don't think it is a groundbreaking revelation to make a higher drag vehicle (e.g. Toyota Tacoma) to have a rear curvature as a function of height that matches a respected lower drag design (e.g. Toyota Prius).

Just like copying another engine's air/fuel ratio. It works on that small turbo four so it's sure to be best for my large naturally aspirated V8...

But I also don't think that measured long term tank to tank 10% better MPG is the "wrong track." Perhaps it is a "non-ideal track", and there is perhaps a percent or so left on the table.

How would you know if you've not tested alternatives?

If I was a professional, with all the professional modeling resources at my fingertips I could know for sure about that percent.

Professionals don't use a template - and barely even mention them in their textbooks - because templates are of such little consequence. If they were the fundamentals some people here try to pretend then the textbooks would be full of them. And they aren't.

[aardvarcus]

Julian,

I truly have not noticed the prevalent use of that template here.

In a hemisphere, the top profile matches the side profile, somewhat by definition. Creating that would obviously be different than creating a profile only used for the top or the side. Hence why I do not understand your use of it. In the sense of separately formed top and side profiles, yes the vehicles I am copying use separately formed top and side profiles just like mine, not the same curve. In the sense of matching exactly, no not all of them in all dimensions.

I do not understand the constant relation of aerodynamic principals to engine air to fuel ratios. Air to fuel ratios could be related to my toaster. I set my toaster on 4, and it burns the bread. I don't know if your toaster reacts the same way. But I don't know what that has to do with AFR. And I don't know what AFR has to do with Aerodynamics in this context. I also don't think my diesel in my current project has strong opinions about AFR, so I will leave the air to fuel ratio to those who prefer spark ignition over compression.

Is modifying to mimic imparting different physics from designing from scratch?

I did test alternatives, some of which are documented on this site. I built the best one, but obviously as I did not test an infinite number I leave open the possibility of something better. I am not aware of anyone who has designed an add on aero device for a 2nd gen Tacoma that imparts more MPG gain than the final one I built, but I stand ready at any time to be bested by a better design.

I ran some overlays of the vehicles in your anti-template video, using the commonly used here AST-II, and found the results interesting. Five out of six were a perfect match, either in overall slope or the spoiler rising to hit the profile! The Sagitta was the one that did not match.

I find it interesting that 5 of the 6 vehicles used to disprove the template concept actually all follow a "template", one that is posted and commonly used here.

[JulianEdgar]

Originally Posted by aardvarcus

Julian, I do not understand the constant relation of aerodynamic principals to engine air to fuel ratios.

Copying the air/fuel ratio from a completely different engine and assuming that, because it works on that engine, it will work on your car, is exactly the same approach as copying an aerodynamic template.

To put it in a nutshell, it ignores the complex interrelationships and interactions that occur in the real world.

Using rules of thumb is a really crap way of modifying cars, but is unfortunately widely done.

[freebeard]

Reasoning by analogy has it's pitfalls. Analogizing across domains even moreso. Analogies can illustrate but they can't explain.

Even though they might have Bernoulli's principle in common.

[JulianEdgar]

Originally Posted by freebeard

Reasoning by analogy has it's pitfalls. Analogizing across domains even moreso. Analogies can illustrate but they can't explain.

Even though they might have Bernoulli's principle in common.

I think it's a pretty good analogy - engine management tuning and aero modification with a template.

1) Both relate to car modification.

2) Both have deceptive rules of thumb that are widely endorsed.

3) Both have 'scientific optimal' values that, in the real world, aren't.

4) Both have a lot of people who find the subject difficult and so seek out 'easy answers' (that aren't).

5) Both have optimal values that can be found only through measurement and testing.

6) Both these days can be tested on the road with low cost gear.

But as I said, rules of thumb are very common in car modification, so instead of engine management we could easily pick suspension modification, car sound, etc, etc.

[AeroMcAeroFace]

Originally Posted by aardvarcus

Yes there are different "templates" (diagrams) from different sources.

I am not sure who on this site over the past decade has been encouraging the use of the AST-I over the AST-II. I have been on this site about that long, and have not seen any pressure to use the older version. The OP of this thread is just about the only one I know doing overlays using the old version.

Well I have been around here for a bit longer than that, not as a member though, and I don't remember anyone saying "Don't use the template in the tools section, it is wrong" Why is it in the tools section if it is the wrong tool?

Imagine a builder turning up at your house, wearing a tool belt with a hammer in it, but then every time he needs a hammer he goes back to the van to get a different hammer, uses it and puts it back in the van. You would question why he doesn't put the hammer that he uses in the tool belt.

It seems like whatever shape that a car maker uses there is a different template that fits.
Which is the one that is supposed to predict separation?

And if there is one that can predict separation (along the centreline), surely you need a correction factor for boundary layer thickening and transition to turbulent flow? Upstream effects are important and not usually considered.

I have seen so many people using the template in the tools section for their "assessment" of how aerodynamic a car is, rarely is there someone "correcting" them and using a different template to "assess" the car.

[aardvarcus]

Aeromcaeroface,

I honestly have no idea why that diagram is posted there. I don't have any control over the site. All the diagrams that I have for my use from this site I pulled from threads where I could read the context and the intentions of what they were created for.

I do recall when I 1st joined the site one of my 1st threads involve me posting overlay with the older version, and was promptly directed to the newer version.

From the things I was told and researched about the diagrams I have, I don't have a diagram that is used to predict separation. Not what they are for.

I took the time to point you to the diagram that I thought would be most appropriate for your use a few posts up. Mainly because back in the day people did the same for me.

[aerohead]

Originally Posted by Cd

I remember there was a long drawn out, confusing discussion here on how the EV-1 could have a .19 Cd, yet be so far off of the template.
Aerohead, I think you mentioned something about the tumblehome playing a large part in keeping flow attached once the rear glass goes out of the curve of the template.
Did I understand correctly ?

Gary Eaker is one of the people that worked at G.M. on the E.V.-1 aerodynamics.

I have suggested in the past that some of you here might could invite him to the forum as a guest.
No one seemed interested.
I would invite him here myself, but I'm a "no-body" .
Aerodynamics is just a hobby that fascinates me, and I can't even do basic math.

Rather than argue back and forth, how about let's invite some other experts to the forum.

I'd love to have a guest on the forum that could answer some of the questions here.

I have Gary's contact information if anyone is interested.

* According to General Motors, the EV1 was Cd 0.197. As an 'ICE' vehicle, we'd be looking at around Cd 0.232.
* When I did the initial profile comparison, all I had was the AST-1, which is probably best suited to shooting-brake type vehicles of long rooflines.
* Out of respect for Julian Edgar I'm re-visiting some vehicles with eight different templates. Just using the AST-II makes a remarkable difference in interpretation.
* I've not yet done a second dimensional analysis of EV1.
-------------------------------------------------------------------------------------
* Gary Eaker would be great to bring onboard. I've planned to write him for a long time now, and hope to complete the Spirit of EcoModder project one day at the A2 wind tunnel.
* A peer-review jury would be awesome!
* I did investigate the HOT ROD Magazine Camaro land speed record car, which I believe Gary Eaker consulted on. The 'Bonneville' rear spoiler on that car happens to match the AST-II perfectly.

[aerohead]

Originally Posted by AeroMcAeroFace

Attachment 29641

I must have done this wrong then. I have tried quite a few times though. I think template wisdom would suggest reaching up to the template to get lower drag.

They must have separated flow all the way down the back.

"Aerohead, I think you mentioned something about the tumblehome playing a large part in keeping flow attached once the rear glass goes out of the curve of the template." Flow is 3d, there are C-pillar vortices that can cause the flow to attach. However it is not always the result of C-pillar vortices, the EV1 had a lot of attached flow from the side of the car going over the back window. I would think that the XL1 is the same.

I don't have an explanation. I used the AST-II. It's dead nuts on.

[aerohead]

Originally Posted by JulianEdgar

Only Aerohead would think that he knows more about the content of a book than the person who wrote it.

Unbelievable...

But that's the very man who people have apparently believed here over all those years. It's a funny world.

A critical reading of the text revealed a logic discontinuity ( reversal ) between page -186, and page -195. It's one of the few sticking points I observed in the book, however of the highest magnitude of importance, with respect to the audience, and their capacity to resolve the conflict in thinking.
That's all I'm saying.