Designing a tail extension for the first-gen Insight

[MetroMPG]

Re: the length of my cardboard boat tail:

The tail as you see it only adds roughly 4.5 feet to the overall length of the car, beyond the OEM bumper, including the tail lights glommed on the tail's transom.

The 62.5 in. measurement you see in the diagram was just there to provide a reference to another Metro owner who inquired - it shows the length from the forward edge of the rear taillight assembly to the tip of the tail.

There's still a fair amount of "car" underneath the forward end of my boat tail as it appears in profile.

[RobertSmalls]

Originally Posted by aerohead

Jim,if you can compare the area at the back of your new tail to the area of the back of the Insight,it will give you an idea of your new Cd.The profile drag varies directly with wake area.

True, but we don't (and maybe can't) know the contributions of profile drag vs skin friction, nor how much of the profile drag happens at the rear of the vehicle.

Come to think of it, maybe you can estimate skin friction's contribution to Cd, by looking at the Cd of a flat plate in turbulent flow. That would give Cd=0.03 from skin friction, for any object that is roughly shaped like a compact car, assuming 100% turbulent (and perhaps also 0% detached) flow.

If a tail increases surface area by a third, you'd be looking at Cd=0.04 from skin drag, and the rest is form drag. But don't expect to reduce profile drag to zero by reducing transom area to zero. Basjoos achieved Cd=0.17 with a tail and other mods, and any car shaped just like GM's race prepped Impact with tailcone will have a Cd=0.153.

Darin, did you record before and after transom area and CdA on your Metro?

[MetroMPG]

I didn't try to measure the "after" Cd.

But you can get the transom area of the boat tail from this photo:



I'll take a photo tomorrow to estimate the transom area of the car without the tail.

[MetroMPG]

Also, I doubt you're going to see any reversed flow on the end of the hatchback (upper) glass. Did you get a chance to try it (tuft testing)?

The smoke trace shows boundary thickening, but I can't see how there would be separation/reversal/recirculation there.

Note also the smoke trace at the front of the car doesn't start on the surface of the hood, so one would expect a gap at the end as well.

[silverinsight2]

Originally Posted by MetroMPG

Also, I doubt you're going to see any reversed flow on the end of the hatchback (upper) glass. Did you get a chance to try it (tuft testing)?

The smoke trace shows boundary thickening, but I can't see how there would be separation/reversal/recirculation there.

There is energized airflow here. The whole length of my rear window stays clear of rain at speed.



Five AirTab vortex generators were placed across the trailing edge of the rear hatch.

After fitting the vortex generators, the measured trip fuel consumption immediately rose to 3.0 - 3.1 litres/100 km, a 3-7 per cent increase over the 2.9 litres/100km average. No change in the feel of the car could be felt - there was no apparent improvement in stability, reduction in wake noise or any other positives. Simply put, the vortex generators made things worse.

Julian Edgar


Also a little information about Insight bellypans and rear tire airdams.
autospeed.com

[RobertSmalls]

I tuft-tested today. The angles of the stock glass are fine. Yarn mostly lays straight back at 64mph, but the tufts near the left and right edges of the hatch glass dance around more than the others. The mirrors might be partially to blame for this.

Originally Posted by silverinsight2

Also a little information about Insight bellypans and rear tire airdams.
Browser Warning

I will build something similar, but a fairing BEHIND the tire would have a larger effect.

[orange4boy]

I agree that behind the wheels would be better. I think those wedges have a few problems. One is they would create vorticies on either side if they were parallel to the airstream. Two is they are probably not parallel to the airstream. Three, they probably create a restriction, forcing more air between the tires but not in a good way.

[aerohead]

Originally Posted by RobertSmalls

True, but we don't (and maybe can't) know the contributions of profile drag vs skin friction, nor how much of the profile drag happens at the rear of the vehicle.

Come to think of it, maybe you can estimate skin friction's contribution to Cd, by looking at the Cd of a flat plate in turbulent flow. That would give Cd=0.03 from skin friction, for any object that is roughly shaped like a compact car, assuming 100% turbulent (and perhaps also 0% detached) flow.

If a tail increases surface area by a third, you'd be looking at Cd=0.04 from skin drag, and the rest is form drag. But don't expect to reduce profile drag to zero by reducing transom area to zero. Basjoos achieved Cd=0.17 with a tail and other mods, and any car shaped just like GM's race prepped Impact with tailcone will have a Cd=0.153.

Darin, did you record before and after transom area and CdA on your Metro?

No doubt things have changed since the 1970s.Back then,profile drag constituted 55% of drag,skin friction was 7-12%,the remainder locked up in cooling sys. drag,interference drag, and induced drag.
With respect to the rear,if you're finished modifying the front,then the rear is where the rest of the drag is.The only way you can eliminate it is with a full boat tail.If you reduce the transom to zero you will reduce profile drag to zero if you stay on the template.
With respect to basjoos car,it can still go 0.08 lower.
The EV-1 could go 0.063 lower.

[3-Wheeler]

Originally Posted by aerohead

...With respect to basjoos car,it can still go 0.08 lower....

AeroHead,

Now I'm curious...

Would you care to elaborate on what you would do?

Thanks, Jim.

[aerohead]

Originally Posted by 3-Wheeler

AeroHead,

Now I'm curious...

Would you care to elaborate on what you would do?

Thanks, Jim.

Jim,this was one of my quirky thought experiments and I hope basjoos won't be offended.This is what I did.
From a side shot that basjoos provided I drew in a ground plain under the tires.With a French-curve I extended the apparent hatch line rearward and projected a prolate ellipse forward to capture the nose of the Civic to ground level.
This established a 'template' around the car,as Jaray,Kamm,Lay,and others had done with development vehicles.
Locating the 'top' of the roof camber I was able to establish the position of frontal area at 44% of overall body length,aftbody constituting 56% of body length and L/D= 2.416.Exit angle at zero ground clearance is 35-degrees,and the exit angle at the boat tail 'hatch' is 26-degrees.
If I stretched only the aft-body out to 1.78-D ( D is 2H ) which is the aft-body of the 2.5:1 streamline body,the new point at ground level to define the 'effective' tail would produce Mair's 22-degree angle.
The intersection of this new template with the line of the undercarriage projected straight back,would define a body with Cd 0.12 for a standard Civic.
NOTE: The diffuser portion would not respect SAE's 10-degree break-over angle and this part would have to be 'active',or the tail unit would have to comprise a trailer rather than a full cantilever tail.
I've brought nothing new of my own here.Kamm's full tail produced Cd 0.15.Jaray's,with a front airdam got Cd 0.13.Walter E.Lay produced three cars with Cd 0.12 this way.AeroVironment got Cd 0.089 with wheel fairings on the Sunraycer,Honda Dream-2 got Cd 0.10.
I believe that a convex windshield,careful cooling system ducting,fairing of the tires and the tail mods would push the Cd to,or below 0.10.
Practical?