Belly Pan Build Thread - Insight Gen 1
 

I've almost started this project several times. Could not see how to get past a few barriers. Now I'm ready to address it again. I want to have a smooth belly pan, front to rear. As smooth as I can get it, that is.

Much credit is due to Ecky and Natalya who have shared photos and notes on their belly pan builds. And to Scott, who makes and sells the excellent aluminum front belly pan replacement that I and many other Insight Gen 1 owners are using.

My exhaust hangs down a bit below the level of the oem rear belly pans. I'm planning on lengthwise spacer strips on either side of exhaust tunnel, to position the panels about 1.25" lower, for clearance. Details later. Anyway, hopefully no belly pan cutouts unless the exhaust system needs cooling air flow.

The car is now lifted for access to the underside. There's more to this than first meets the eye. Put the front up on jack stands first. Lifting the rear first caused the front to angle down too close to ground. No access to see or get a jack in there. The rear wheels are on ramps to take advantage of the hand brake to prevent rolling off. The ramps point away from car so I have room to slide underneath at the sides. So - you can't drive the car up onto the ramps in this configuration. A jack goes under the rear coil spring support to get the wheels up, then slide the ramps under the wheels.
https://lh3.googleusercontent.com/zj...=w1024-h498-no

Really glad to see you back in this, Brucepick. You were a big presence with the HX Civic back when I started on EM seven years ago. My own bellypan dreams foundered too, for similar reasons. And I did not have a lift! But I did build the smooth diffuser.

Cali98, what's a smooth diffuser?

I'm starting at the front. Made some progress today.

My car has one of Scott's aluminum front belly pans. Like the oem panels, it attaches to the oem support bracket at its rear, about even with the front wheels' centers. My first sub-project is to build a section beginning where Scott's panel ends.

Here's a drawing showing a cross section of the rear portion of the aluminum panel (blue) with the oem support bracket (black). I added a 1" x 1/8" x 25" backing strip going across the top side of the panel. That will give the fasteners for the coroplast something more substantial to anchor to.
https://lh3.googleusercontent.com/lM...=w1024-h330-no

I considered resting the coroplast on top of the bracket. That is, the horizontal cross member that supports the trailing edge of the forward belly pan. But the up and down features in the aftermarket panel and also the bracket, aren't ideal for aerodynamics - so I plan to hang the new section to the underside of the aluminum instead. So covering up those irregularities.

I will use rubber well nuts to secure the leading edge of the coroplast to the underside of the aluminum panel. They work a bit like expanding wall anchors. Tightening a screw into them forces them to expand, so they won't pull out. The screw is removable and so is the well nut. Mine take #10x32 screws, need a 3/8" hole.

I figured the aluminum belly pan by itself might not be enough for the well nuts so I added the 1/8" backing strip. I used seven 1/8" diameter pop rivets to secure the 1/8" bar stock backing strip above the panel. Rivets inserted from the bottom side.

Here's a drawing showing a compressed well nut with washer and screw, and a pop rivet. Shows the belly pan (blue) and the 1/8" backing piece. I omitted the layer of coroplast that will go between the well nut and the washer, sorry.

https://lh3.googleusercontent.com/jv...=w1024-h323-no

Here's a photo of a 3/8" hole with well nut partially inserted. The panel is upside down here. The well nuts and their screws will be inserted from the underside when assembling it.
https://lh3.googleusercontent.com/_5...=w1024-h767-no

Here's a photo, panel still upside down, with well nut inserted and a screw + some washers tightened into it. All stainless because I hate when fasteners get rusty. Tightening the screw squishes the well nut so it expands; it won't pull out.
https://lh3.googleusercontent.com/Jv...=w1024-h767-no

Angled properly but without strakes. :thumbup:

Yes, thanks. I'm aiming for that also. Due to the insight's configuration back there, exoect little upward slope as the lowernaero pan meets the bumper. But I will see when I get that far in the project.

This morning it snowed here, with rain forecast for tomorrow and Wednesday. Work is on hold till thendriveway is dry agin. Hopefully on Thursday..

Remember that you don't need much rise at all. Calculate your angle. You might find you have too much rise, axle to rear bumper.

http://ecomodder.com/forum/member-ca...rding-body.jpg

It took me a while to figure out that graph. Soooo...

Since my Insight is an aerodynamically shaped hatchback I use curve "a".
If my diffuser has a slope of 7.5° compared to the previous section's slope, the diffuser would have zero effect on overall Cd. If the slope is 10° different, Cd goes up by about .008. Whew.

Is that right?

Yup. Three degrees is .02 reduction in Cd.

Thanks for this. Good news.
I did the calcs for this, for my car and I like the results.

I get an angle of 3.69º for the diffuser, using existing available attachment surfaces. That's assuming the panel prior to the diffuser is horizontal, which I believe it is. Pretty close, anyway. The rear bumper lip is 2" higher than the existing support I'll use to support the diffuser's leading edge (which will also support the prior panel's trailing edge). The diffuser is 31" long at the center, so that's an angle of 3.69º per a handy calculater on the web. Allowing for the bumper's curve, if the diffuser length is only 25" at the sides, the angle is still 4.57º. According to the chart you posted, either case gives a reduction in Cd. Thanks!

Thanks for this. Good news.
I did the calcs for this for my car, and I like the results.

I get an angle of 3.69º for the diffuser's upward slope, using existing available attachment surfaces. That's assuming the panel prior to the diffuser is horizontal, which I believe it is. Pretty close, anyway. The rear bumper lip is 2" higher than the existing support I'll use to support the diffuser's leading edge (which will also support the prior panel's trailing edge). The diffuser is 31" long at the center, so that's an angle of 3.69º per a handy calculater on the web. Allowing for the bumper's curve, if the diffuser length is only 25" at the sides, the angle is still 4.57º. According to the chart you posted, either case gives a reduction in Cd. Thanks!

The calculator I used

I did some other measurements today, on the exhaust system's low points. They are 2" below the existing lip where I will attach the center panels' outer edges. Based on where I can place lengthwise runners to "press" the panels lower - to clear the exhaust low points - that gives an angle of 4.86º at those added bend points. But I'd need to extend lower than that to have true clearance below the low points, for protection against heat and vibration.

So an additional 3/4" of clearance below the low points, would create lengthwise bends in the panels of 6.67º. I might rather keep the panels relatively flat, and make a couple cutouts for the exhaust protrusions.

This is a poblem that discouraged me too. (And my desire not to drill the body.) Nonetheless, a 6.67 angle is better than a 90 degree angle any day, no?

I'm afraid the lenghwise runners - to force the panel down below the exhaust near the center line - would probably becomes a case of what my Dad called in Yiddish "tzimmes machen". Basically, it indicates a huge amount of highly detailed work or analysis, to the point where you can't tolerate doing it. I'm leaning towards flat panels with simple cutouts for the exhaust low points.

The rails to drop the panels' center line would create their own protrusion, and would require a good bit of effort to fasten the rails to the underbody and then to fasten the panels securely to the rail. I think T-nuts would work with something like a 2x2 (which is actually smaller than 1-1/2" square), but it would still be a pain to build.

Tzimmes is a Jewish baked sweet dish. Lots of cutting and a two-stage cooking process. Typical recipes could include some or all of these: carrots, raisins, plums, celery, onion, potatoes, sweet potatoes. All of these cut up into small bits and stewed together, with sweeteners such as cinnamon, honey, brown sugar. After stewing, bake in a pan. Not a quick process, usually.

Today I did a bunch of prep and reconnaissance under the car.
It was pretty cold - with wind chill, it felt like 35º F. Progress should pick up when the weather gets warmer, hopefully soon!

I measured the positions of the added fastener positions in the aluminum front panel, and marked them on the underside of the panel. So when it's back in I'll be able to measure out the locations for the holes in the first coroplast sheet. And I reinstalled it. All the well nuts are in, held in place with screws and washers. The coroplast panel isn't ready yet.
https://lh3.googleusercontent.com/1Z...=w1086-h815-no

I got both oem plastic aero/protective side panels off, and removed the fasteners along the rocker panel bottoms. So the new panels can get tucked under the rocker panels' inner lips.

Those fasteners like to break on removal; they're old. I maybe can get similar replacements. Or since their holes are 10 mm diameter, (rocker panel mount holes) I can use well nuts made for #10 screws, similar to the #8 well nuts I used in the fastening points I added to the aluminum belly pan.

As Ecky and Natalya did, I'm planning on spanning the sheets across the car. The joints between pieces will run crosswise. I plan to fish scale the joints - overlapping in the direction so the air can flow over it. I'm not going to try for butt joints.

I also made measurements for the first (forward) coroplast panel.
It needs to be 46" wide by 24" front-to-rear. That will reach from the row of mount points I added to Scott's belly pan, going back to the first four mount points originally used for the oem panels. Those are M6 screws, with 10 mm heads. It won't cover the full width of the car at the firewall but that's OK.

I have 3 coroplast sheets, 48" x 96".
I will cut one down to 48 x 70.
That leaves a piece 48 x 26.
That will be my front piece, that needed to be 46" x 24". A couple extra inches will be fine. I doubt it will matter which direction the coroplast channels run.

As for the remaining piece - 48 x 70:
In the area between the firewall and the rear bumper, the maximum width I found that I'll need is 68". That's at the firewall, it tapers narrower as it goes back. So the 70" wide piece I'll make is planned to be able to span that width.

I've numbered the panels for reference in this thread.
All panels are overlapped (front one below the rear) like shingles or fish scales. The overlap joining #2 and #3 is large, due to the locations of the existing available attachment points. They're not in an even line going across the underbody.

Panel 1: The oem front panel section. Aluminum + rubbery plastic.
- Or in my case, the aftermarket aluminum replacement panel, by Scott. AKA klr3cyl on Insight Central. It replaces the aluminum plus rubbery plastic oem front section.
Panel 2: Goes from the rear of #1, to about a foot rear of the firewall.
Panel 3: Goes from the rear of Panel 2 (a foot rear of the firewall), to a point about a foot forward of the rear wheel wells. About 48" lengthwise.
Panel 4: Goes from the rear of Panel 3 (about a foot forward of rear wheel well), to the crosswise narrow air deflector that's mounted at the rear of the gas tank. Air deflector = attachment point.
Panel 5: Goes from the air deflector at rear of gas tank, to the rear bumper lip.

Nice work, Bruce. Lots of good homework, measuring, and thoughtful installation work. Me like.

I have more posts coming.
I just finished cataloging 22 photos for this thread.

I finished off this project during the seven days April 5-12, just in time to drive up to the Green Grand Prix. I took a number of photos but had little time until this weekend, to catalog them and upload them to Google Photos for use in this thread. I'll come back to this again as time permits.

I'll have some details on such captivating topics such as attachment, sag prevention, and aero smoothing of the joints between panels. I leaned towards methods that would not be too time consuming but should give a sturdy and aerodynamically clean result. I don't claim to have done the best job possible but I think it's a good job.