Truck aero cap
 

I have been wanted to get an aero cap for my truck for a while now. I used the ecomodder tool to get an idea of what it should look like.

Attachment 33576

.Here's the proof of concept from xps foam, glue and paint. I hope to use what I learned from this versions and build a better one:

Attachment 33577

Tuft test looks pretty good for a first crack.
Tuft test Video:
https://photos.app.goo.gl/Pp9tYJTwWnyYEpcbA

Hope to get A-B-A testing on version 2 as the weather warms up in Michigan.

What construction technique for the final version? Matching or contrasting finish? I'd consider a cedar strip canoe style construction

Else shingles laid horizontally, like fish scales.

I am planning on fiberglass for version 2. Also designing in a "hatch back" style that can open. It also has to be easier to put on and remove as I use the back of my truck frequently.

A fastback hatckback is a tall order. Posibly have the back half or third hinged completely.

On my 2004 Tacoma, I did something very similar. I was not looking for max mileage over functionality. Regardless the results surprised me. I wanted to retain rear window useful visibility via the center mirror, and I wanted the rear sliding cab window to remain accessible, for the occasional transport of say a long pipe. Built a 2x10 sided wooden box, with a top flat section of 1/2" Baltic birch PW approx 60" wide x approx 42" long, At the 42" point I straight tapered the 2x12's down to approx 1"h, and over hanging the tailgate approx 3", and then hinged another piece of Baltic birch PW. This overhang had a 2x2 near the end L to R which effectively locked the tail gate from being opened without lifting this rear hatch. I made removable legs so I could securely prop up level the rear hinge hatch to haul items as needed on the 84" combined flat surface, which I reinforced to handle loads I expected to carry. I installed locking pins to secure the rear hatch, I turnbuckled the entire assembly to inside the pickup bed for easy removal.
This truck has currently 550,000 miles on it, it has a long record of expected fuel mileage. First trip with the box/hatch assembly of 3,000 miles, mileage increased from 22 to approx 23.8, even with the approx 100+lbs of extra weight. When I returned, I did a wool test to try understand what was really going on, and found out just by luck/gut, I had the hinge point/taper start right where it seems airflow began to reattach after flowing over truck cab/roof. Not sure what part the tailgate hatch overhang/lip played in this entire experiment. And to enhance usefulness, on both sides I attached externally to the 2x10, I secured alum aero tie down snap rails to increase flexibility.

'tuft testing'
 

Flow looks good!
It does indicate the formation of attached longitudinal vortices, along the roof / sail panel intersection of the cap, which carry relatively high drag.
A solution would be either chamfering, or softening with a radius matching that of the cab.
Radius is best, but also the most intensive, as far as fabrication goes.
-------------------------------------------------------------------------------------
Also, you're creating a 'fastback' roofline, which is the most sensitive of all roof designs to tumblehome.
If you stand at the back of a HONDA Fit, looking forwards, you'll notice that the A-pillar, B-pillar, and C-pillar are all at different tumblehome inclinations.
The airflow really 'likes' this.
Again, these subtle features are the 'most expensive' to fabricate.

This is the area you are referring to?
Attachment 33586

It appears the air pressure is lower on the roof vs the side since the upper tufts are getting pulled up and over the edge. This edge will be difficult to round. At best it will be between 1-2 inch radius. However the angle as it moves back does reduce. It is hard to see but the angle starts at about 75 degrees and ends around 45 degrees. The roof cross car dimension is consistently 48 inches. So with the edge following the truck side, as the roof comes down, the angle changes to accommodate. Here's a view of the recent CAD:
Attachment 33587

As for rear visablity, I and sticking with side mirrors.

I question the advantage of increasing the radius in that area, and wonder if more might be gained by adding a small aero fence each side to prevent the side air flow migration up and over to what appears to be almost laminar airflow on the upper surface.

I actually had that thought also. Sort of feels like the winglets on the end of a wing to reduce the area of high and low pressure.

There is a youtube video out there by Julian Edgar where he did that on the back of a car. He has tested the rear finns on a honda insight.

'referring to?'
 

Yes, exactly. It's the Bournoulli Theorem at play.
Air over the truck is taking a longer path than along the side, yet arriving at the same time; requiring a higher velocity to compensate.
There's a direct trade-off between velocity and pressure, leading to the lower pressure up top, which attracts the slower, higher-pressure flow alongside.
--------------------------------------------------------------------------------------
Without radii at the intersection between the top of the new sail panel and roof, we lose the ability for some comingling of the two streams and the shearing forces they pose.
FIAT published a 'with' radii, and 'without' radii, fastback roofline comparison in an 1986 SAE Paper, and found up to a Cd 0.035 drag penalty for the roof lacking the edge softening.
Member 'bondo's' Aerolid aero shell was wind tunnel tested, courtesy of the Ford Motor Company, and returned a Cd 0.067 drag reduction on a 2012 F-150.
The FIAT data infer that, without the cab-matching radii on the aero shell, the drag reduction could be reduced to only delta-Cd 0.032, a 52% loss in performance.
------------------------------------------------------------------------------------
This vortex-drag was studied intensively by 'Kamm' at the FKFS.
There are no low-drag cars ever fabricated which did not possess this continuity between the forebody roof edges and aft-body.
--------------------------------------------------------------------------------------
Skipping ahead to adding 'fences':
* On an aircraft, like a Boeing 737, operating at 33,000-feet, and for a perfectly-streamlined wing, in two-dimensional flow, the wingtip treatment would be a must.
* On a non-streamlined, bluff-bodied, road vehicle ,operating within three-dimensional flow, and ground proximity, adding some sort of 'winglet' would be a violation of the ground rules for fluid mechanics, as spelled out by Dr. Hucho, at least in his December 1987 2nd-Edition ' Aerodynamics of Road Vehicles.'
These fences have Cd 1.15, and in the statistical 7-mph crosswind that the truck is always go to experience, you'd be installing a 'trigger' for span-wise contamination, and introduce a roll moment during wind gust.
------------------------------------------------------------------------------------
' the vehicle aerodynamicist must refer to a large amount of detail resulting from earlier development work.' Hucho, PREFACE, 2nd-Edition, 1987.
--------------------------------------------------------------------------------------
My Toyota T-100 uses a 69mm ( 2.75-inch ) radius to marry the aero shell's roof and sail panels together, mimicking the T-100's cabin.
You're not 'adding' radius, you're just providing a missing ingredient.