Roof Rack for Honda Insight

To my knowledge, there is only one commercial roof rack available for the Honda Insight, and I've read that it requires care to avoid roof damage.  But I wanted to be able to carry a canoe, or other big things, such as lumber or a ladder.  So I built the rack described below.  Some might doubt whether the Insight is suited to carrying large cargo like this, but in fact, it is overkill--canoes are light and can be, be example, towed behind a bicycle.  The rack does drastically degrade the excellent aerodynamics of the Insight, particularly with a canoe, but for moderate speeds, it works fine. 

Canoe on Honda Insight on custom wooden rack

The rack is made of wood bent to fit the curve of the Insight's roof.  With rubber padding, it distributes the weight over a wide area of the roof, all directly above or adjacent to the sturdy beams that are designed to provide rollover protection.  So it should be able to take lots of weight--much more than a canoe.  It is held in place by two straps run through the doors (and are typically also used to hold the cargo to the rack), by the rubber padding that provides high friction against the roof, and by the match between the curvature of the roof and the rack.  This relatively temporary mounting scheme is appropriate since one wouldn't want to leave it on the roof with no cargo and hurt the aerodynamics unnecessarily.

This page summarizes the construction and my experience using it.

Wood roof rack on Honda Insight

Construction

I considered three ways of making wood bent to fit the curve of the roof:  1. Cutting curves into large pieces of wood, 2. steam-bending sturdy pieces of wood, or 3. bending strips of wood thin enough to bend without steam, and holding them in their bent position with larger pieces of wood.  I chose the third option because it seemed the easiest way to match the curvature, and because I'm not set up for steaming large pieces of wood.

The materials are: 2x3 (nominal) studs for the two sides, pine handrail stock for the cross rails, and pine trim strips bent to match the curvature of the roof.  I used handrail stock because it was one of the few shapes that Home Depot had in FSC certified wood; the bent trim strips are also FSC certified.

The construction proceeded as follows:
I would have hoped to be done at that point, but I'd actually never checked whether the width, designed to match the car, would match the boat.  It was too narrow in the back.  So the canoe, rather than sitting on the cross bars (handrail stock), sat on top of the 2x3s, rather precariously.  To secure it, I had to cut two more 2x3s and attach them to the back sides to hold the canoe in place there (below, back left).  And I cut two new blocks for the front to hold it level (below right).  Finally, I also cut two more small blocks with two prongs each to hold the center thwart of the canoe (below center).  The second picture below shows all of these holding the canoe.
wood to hold canoe in place

wood hooks holding canoe thwart

To hold the mortise and tenon joints together, I used a hook and eye, but I used a machine-screw thread on one eye, backed by a wing-nut, so I could tighten it, and I used a latching link as a hook.  I plan to put a cap nut over the exposed end of the thread, to avoid poking anyone with it, and to make sure I don't lose the wingnut when it's loose.
hood-and-eye to hold rack together   wingnut for tightening hook and eye

Finally I needed to fasten everything in place.  I originally planned on strapping the rack to the roof; then the canoe to the rack, in order to make it easier to get the canoe on without bumping the rack off.  But so far that hasn't been a problem, so I might just stick with using one or two straps around the whole thing as shown below, plus back and front straps, as shown below and in the first photo.   The rear straps connect to the cargo net hooks inside; the hatch is closed on them, and then they are tightened.  The front could use the tow hook points for a very secure mounting point, but it seemed fine to simply loop around the bumper twice

canoe on homemade wooden rack on Honda Insight

Use experience

I just finished this and tried it on just one trip to a nearby lake (about 10 or 15 miles each way).  It seemed very stable; much more than the foam-block approach I've previously used with canoes on various cars.  No trouble at all with it shifting, despite bouncing around with 45 psi tires on a rutted dirt road.  It's hard to judge its effect on fuel economy, as the trip was all on 30 to 40 mph roads.  I got about 65 mpg on the way up, which I thought was pretty low, compared to the  82 mpg average I've had since owning the vehicle.  But then I got 104 mpg on the way back down, so apparently the uphill on the way to the pond was more of a factor than the aerodynamics.  But I could definitely tell that there was more drag, based on how easy it was to maintain speed in lean-burn, how far it would coast, etc.  And that was at relatively low speeds--I didn't try going above 40 mph.