35. Splitting h-air-s

When I spec’d out my kit I had a really difficult time deciding between the street splitter versus the track splitter.  Ironically, the higher performance track splitter is actually cheaper than the street splitter.  This had me scratching my head for a bit but I eventually discovered the track splitter comes as a “skin” whereas the street splitter is ready to use as-received (kind of).  The track splitter “skin” is a thin shell of the splitter, it’s up to the builder to make the “guts” of the splitter so it’s solid.  The street splitter is made up of a fiberglass honeycomb core sandwiched between sheets of fiberglass.  It’s solid and can be installed and used without further modification – that is, unless you’re as worried about front end aerodynamics as I am.

The real reason I didn’t go with the track splitter is it lowers front end ground clearance by 1-2″.  For my street car, and on the steep grades around my neighborhood, that ground clearance is critical.  At the time of ordering I wasn’t aware the track splitter needed so much modification to make useful – I’d have been a bit frustrated getting the part and realizing this after the fact.  So here’s your warning future builders!

Of course the path I went down required a fair bit of modification of the street splitter anyway.

I added the splitter tunnels for my street splitter.  The tunnels are large expansion chambers you add to the splitter to create a venturi effect under the nose of the car.  This in turn creates additional downforce.  On a car with a big wing on the back end, more front end downforce is critical for maintaining aero balance of the car.

Here’s what I did to install the splitter tunnels:

img_0947
As-received, the tunnels are not flat.  These need to be bonded onto the splitter and having these large gaps will make it difficult to get a good bond.
img_1028
Some resin, cab-o-sil, and chopped fiberglass and we’ve got a flat tunnel!
img_1094
I added holes throughout the flange to give the panel bond adhesive more surface area to grab onto.
img_1092
Similarly, I added holes to the splitter.  It’s a pretty big cut out for the tunnels!
img_1108
A close-up showing the internal honeycomb structure.  Looks like a bunch of fiberglass cylinders, pretty cool.  The outer edges are filled with resin for impact resistance.
img_1095
I added some temporary fasteners to keep the two parts together while the adhesive was curing.  I didn’t think about it at the time, but had I been more intelligent about where I made my holes I could have left the bolts in place permanently.  I’m going for a belt and suspenders solution here (adhesive + fiberglass + fasteners) to keep the tunnels in place!
img_1106.jpg
Tunnels bonded.  The transition’s not very smooth at this point.
img_1111
Reinforced resin to fill all the holes and build up the transition at the front of the tunnel.
img_1112
I chamfered the inlet somewhat to try and make it a better transition for the air.  In truth I’m not really sure how well these tunnels work.  RCR states these tunnels make the street splitter almost as effective as the track splitter.  Generally a tunnel such as these should be gently introducing air into the expansion chamber.  The steepness and height of these tunnels makes me think there’s going to be a whole lot of flow separation happening in here – not so good for airflow.  But the chamber’s so big maybe it creates a layer of separated, disturbed flow along the walls and a more laminar flow closer to the ground.  I know enough to question but not enough to be certain one way or the other.  (shrugs)
img_1114
Some more fiberglass for added insurance.
img_1123
2 tunnels bonded!
img_1124
What it looks like from the top.
p1010431
Here’s a pic of the track splitter for comparison.  The tunnels are much wider and the lower-most “shelf” of the splitter would actually sit below the plane of the street splitter if they were laid on top of each other.  That middle “shelf” of the track splitter is on the same plane as the street splitter.  What I like about the track splitter is these tunnels have a larger area for the air to move into, potentially creating a smoother flow field.  What I don’t like so much is the tunnels are super wide and half of the discharge dumps directly into the tire – which is a high pressure zone.  This makes it more difficult for the air inside the splitter to exit cleanly.  The street splitter doesn’t have this issue if the tunnels are installed as I have done – toward the inside, and away from the tire.

If there are any motorsport/aero engineers out there reading this I’d be interested in getting a lesson!

I hemmed and hawed for a very long time about whether I wanted to install a hinge for my front clam.  A hinge would make it MUCH easier to service the front end of the car (hydraulic, brake, clutch fluid reservoirs, lift pump, brake residual valves).  It would also add more complexity – not only just designing/packaging a hinge, but also to the layout of everything inside the front clam.

One of the complicating issues is with the front wheel well liner.  The liner closely hugs the front tire and sits squarely in the way of where you want to install the outboard splitter support rod.  Generally this rod attaches near the rear of the radiator box and extends to the front or rear corner of the splitter.  If running the front wheel liner, the arc it swings as the clam hinges forward cuts right through the middle of the liner – the liner would have to lose about 50% surface area down the middle for the support rod to exist.  If, on the other hand, the support rod attachment point were pushed forward, to the middle of the radiator box, the lower 50% of the liner would have to be removed – also not acceptable.

I prioritized keeping the liner intact and figured I would very rarely be servicing the components up front – so I finally decided to make the front clam fixed.  Once I made this decision, laying out the splitter and front end hardware became much easier/clearer.  As is typical, the rear area of the front clam will be secured to the spider via aerocatch latches.  At the front, I’ll be using 2 quick-release spring loaded pins.

Aerocatch latches:

The “nose latch” is typically located just inboard of the upper surface for the front wheel vent.  This location keeps the latch hidden and is the normal location most builders choose to use.  I opted to go down a less traveled path (no surprise!).

My issue with the typical location is the clamping force is applied along the exterior of the body only.  If your bodywork is perfect (or you spent a lot of time shaping the front) this is fine.  However, I had a few issues with my body that necessitated a different solution.

With the front resting on the splitter the rear edges of my front clam wouldn’t seat and align with my spider.  Using pressure along the exterior edge only, I could get the body to close and seat against the alignment pins – cool.  But the issue is the interior edge which meets the spider/windshield area remains proud and there’s a visible step down toward the spider.

img_1125
Exterior edge aligned – note the step down at the gap with the spider/windshield area.  The door is not final fitted yet, but resting on the spider to help me visualize the joints.
img_1126
With some pressure placed near the center of the panel I was able to get the interior and exterior edges correctly aligned.
img_1338
Another shot showing the fits.

I had a more severe misalignment issue on the passenger side.  At first I thought this was due to me hacking the interior so I could create the larger radiator exit – but going back to my as-received photos it looks like this condition existed back then.

img_1127
The right side of my front clam “tips forward”.  Even with the splitter set even and the front nose sitting flush, the rear corner tips forward.  If I apply pressure to the back to simulate the aerocatch the front pulls up from the splitter very slightly.  This will require some extra body working to get right.  For now I’m prioritizing getting the rear properly fitted – giving concessions to the front nose as necessary since a discrepancy there is much less obvious than at the back.
img_1129
A similar condition between exterior/interior gaps exists.

**Miracle of miracles, after I bonded the forward parts of my wheel liners it seems I must have locked the front end in such a way that this “tipping” of the front end has rectified itself!  I still have the fit issue between the interior and exterior, but the outermost gap is now just about dead on.**

img_1330
The gap automagically fixed itself!

If you haven’t figured it out by now, I opted to position my nose latches along the top instead of hidden in the wheel vents.  Placing the latch on the top allows me to center the load – improving the load distribution and helping with my interior/exterior gap issue.  It’s not as nice aesthetically but I’m a poseur and can claim “race car” looks ;).

Positioning the aerocatch took me some thinking – and I may have out-sussed myself.  Initially I wanted to position the latch body centered over the front “beer can holder”.  It’s a ~3.75″ round depression in the front spider – a vestige of an older latch design.  The latch body is too big to fully sit within the cup but with enough of a vertical gap, it may be possible.  Instead of going the easier route I had to go with the more difficult.  Instead, I opted to center the latch pin within the round depression.  This pushes the latch body aft just slightly – enough that it now interferes with the spider :(. In hindsight I think I should have tried centering the latch body with the center of the depression, pushing the latch pin forward just slightly.  (sigh) 20/20.

img_1297
The beer can holder (or Diet Coke in this case).
img_1085
Plotting out the cup holder.  The large square indicates the outer diameter of the depression below.
img_1086
With the center point located, I used the aerocatch template and aligned the body cutout with the pin centered.
img_1100
I crept up on the hole size here – the template is slightly oversized relative to the actual latch body.  I wanted to preserve as much material as possible as the thru-holes for the securing bolts are very close to the cutout.  You can see in this photo where the interference with the latch body occurs – there’s not quite enough room for the tail end of the latch body once the front clam is fully seated against the spider.
img_1116
Lower latch body modified, upper body unmolested.  About 1/4″ of material needs to be removed to create the required clearance.
img_1117
Standard vs modified latch bodies.
img_1118
With the latch handle fully opened it doesn’t dip down far enough to interfere with the spider – phew!

Once I get the bodywork off again I’ll install the latch pin.  My plan is to reinforce the cup holder by bonding a sheet of aluminum onto the underside.  I then plan to bond a nut to that plate so I can make pin height adjustments without having to reach the lower nut.  An upper nut will be placed on the top side – tightening the upper nut will act as a jam nut, securing the pin in place.  If the lower nut isn’t fixed some method of accessing it would be required to make pin height adjustments – a very difficult place to reach unless you cut an access “window” somewhere else in the body.

The forward securing mechanism becomes much easier and cleaner without the hinge requirement.  I bonded and bolted an aluminum plate to the internal vertical fiberglass fins (belt & suspenders!).  On the splitter, I bolted a piece of L-metal that butts against my reinforcement.  I’ll then drill a hole through both pieces of metal and secure them with a quick release pin.  To remove the front clam I would pull the 2 release pins and undo the aerocatches.  I would then need a strong buddy to help me lift the front off the car.  It’s not as elegant as a hinged front but I think this is the best solution for me.

img_1140
Bolts shown are temporary, they’ll be turned around so the button heads are facing outward to preserve ground clearance and mitigate against interference as the clam is being lifted or lowered.  The holes for my quick release pins will be drilled once all the bodywork and front trim are complete to ensure proper alignment.  I’ll likely add a strip of rubber or felt between the splitter and front clam to prevent fretting damage.

Because I recontoured the front edge of my front wheel wells, the factory-located splitter could be pushed forward another ~0.5″ inch.  The new position gives me just enough clearance I shouldn’t have any tire rubs with the suspension at full droop, or while turning the wheel with my front lift engaged.  I think it’s critical the wheels can be turned lock to lock while the lift is engaged as it’s a necessary maneuver when approaching sharp elevation transitions in the road.

img_1130
Splitter pushed forward and re-aligned with new wheel well contour.  I was able to re-use the factory alignment pin location on the splitter; a new mating hole was drilled on the clam.

With the splitter’s fore/aft location defined I could now locate the rear splitter support strut.  I debated pushing the strut’s anchor point on the splitter forward but decided to keep it perpendicular to the radiator box, keeping it as far back as possible without interfering with the front wheel well liner.  Using my hand, I supported the splitter near the front corner then at the rear.  With support at the front, I could get the splitter to make a pretty good knocking sound by lightly tapping it – the rear corner was separating from the body and making a knocking sound as it came back.  With some rubber/felt here it would lessen the sound but I didn’t want the back end of the splitter wagging around, cantilevered in the air.

img_1139
Blue tape along vertical fin shows where a slot will be cut to facilitate clam install/removal.  It will need to be lifted straight up but the car is low so this should be manageable.  The rear splitter support will be aligned with the radiator cross brace I installed in my prior post.

If using the factory-supplied strut parts you’ll want to have 1/4-20 left & right hand taps handy as the provided rods are cut-to-length and not pre-tapped.  I used a standard through-hole tap from McMaster and tapped it as deep as it would go.  It turned out I needed to shorten the supplied eye bolts by about 1″ to achieve the proper amount of thread engagement and adjustability.

img_1276
In the vertical direction, the support strut clears the tunnel.
img_1279
It’s placed directly inline with the support rod used to brace the radiator box.  While the box isn’t triangulated, this is the cleanest approach I could come up with for having an air duct, a support rod, and these outboard struts aligned as best as possible.  As received, the support rod is placed further back, close to the footboy.  The radiator box is pretty stiff back there because it’s bolted to the footboy in multiple places along the back.  Where the box is weakest is about where I’ve got the rod located.

An outline of the wheel well liner can be seen on the interior vertical fin.  I was able to rough fit the liners after a good amount of massaging.  I covered some of that in post 27.  My passenger side liner fit pretty well but the driver side needed a good bit of trimming.  I also had to cut holes for the tunnels and access for the headlights and turn signals.  I originally cut the holes large enough so I could get proper line of sight to my headlight adjustment screws.  Turns out I’ll need to make them even larger so I can get line of sight to the turn signals as well (there’s very little room between the turn signal and liner once installed) – I’ll save that for after the liners are bonded into place.

img_1144
I over-cut the clearance around the tunnels and will clean this area up when working on my headlight closeout panels.
img_1145
Another view.  It was a repetitive process of trying to shove the liner in, trim, rinse/repeat.
img_1133
Rough fit of the driver-side liner.  This liner was too wide to fit until I knocked off about 0.5″ on the exterior edge.  I prioritized keeping the interior geometry figuring this was molded more precisely.
img_1135
Rough fit of passenger side liner.  This liner went in much easier than the driver side.
img_1152
There were fairly sizable and inconsistent gaps all over the mating edges.  To get the liner locked into place I added panel bond adhesive along the upper and interior edges.  To keep the liners stationary while the adhesive was curing I secured the liner using 2 bolts along the top edge.
img_1149
The great thing about fiberglass is it’s relatively simple to patch holes such as these.
img_1157
The gaps are inconsistent – I just want the liner secured well enough it won’t shift while I move the front clam around so I can get better access for fiberglassing.  Working in this super confined space is not so fun – I wish I’d been able to get a photo of me squeezed in between the liner and brake rotor while holding a container of resin, it would have made for an amusing photo!
img_1196
After the panel bond cured but before I pulled the front clam I came back and added some more peanut butter to help keep the liner in place for when I move the clam.  This will get a few layers of chopped strand mat and a layer of fibeglass cloth to finish it up later.
img_1197
At the lower exterior corners I added a small patch of reinforced resin to keep it from shifting.
img_1150
Driver side liner installed.
img_1147
Passenger side liner installed.

The last piece of the front wheel liner puzzle are 2 J-shaped pieces used to block debris from getting throw toward the interior of the spider.  The fit wasn’t too bad – some trimming of the forward edge to mate with the front liner and massaging of the rear should be about all that’s needed.

img_1299
Some trimming necessary.
img_1304
As with the other pieces, don’t expect a flush fit.  As I did with the front pieces, I’ll “tack” these into place using panel bond adhesive and some fasteners, then set everything permanently using reinforced resin to fill all the gaps and follow up with fiberglass mat/cloth.
img_1309
At the back.
img_1305
Side view, looking “into” the wheel well.  There’s a contour misalignment between the wheel well and fender liner.  Best bet is to trim as much as you can from the liner piece, then build up a transition to meet the wheel well.

I’ll attach these last pieces of the front liner once I get my aerocatches installed so I can preload the body and get everything aligned as it should be – I suspect once these pieces are bonded into place the body won’t flex very much.

img_1153
I’m hoping I can make headlight adjustments while the car is sitting level.  Removing the clam for every adjustment would be a serious PITA!
img_1184
An issue with using the Hella 60mm projector housing is the outboard unit sticks out past the profile of the liner.  A small bump will need to be added to the closeout panel to accommodate.
img_1190
I picked up an inexpensive set of LED fog lights from Ebay and set them in my fog light recesses to see what the front end would look like – looks OK.  I seem to have some kind of paired circles theme going on at the front …
img_1211
… and at the back as well.  I didn’t set out to do this on purpose but it’s working for me.
img_1161
I mounted the rear lights once again and spent some more time trying to finesse the openings.  I like the way it looks from directly behind the car …
img_1162
… but from an angle this “eyebrow” look kind of doesn’t work for me.  Oh well, too deep into it to change gears now, it may grow on me – or not, I won’t be standing behind the car much anyway.  A little more bondo work to make the gaps more consistent around the lenses and we’ll finally be finished with the rear taillights!  It was a lot of work to keep the double round look.  In hindsight it may have been better to go with the HJones modification – glass this whole area in and make it flat, then mount 2 (separate) lights in place of the single assembly.  I considered going down that path initially but couldn’t find a round LED light assembly I liked.
img_1208
With the lights mounted and rear wheel well installed, it’s apparent the light assembly sticks too far outside the liner.  Similar to the headlights, the taillight closeout panel will need a few contours added to compensate for this interference.

While on the topic of things I would have done differently and hindsight being what it is, I guess it’s inevitable I’d have to go back and fix some of my earlier mistakes.  The very first fiberglass modification I did was to add a reinforcement patch where the engine’s intake tube protrudes through the rear clam.

When I glassed this in initially I didn’t add anything between the fiberglass and intake tube to create a gap between the two.  This meant the new fiberglass I added would rest against my intake tube at all times – it’s going to vibrate and rub the tube and eventually make it look pretty ugly.  What I should have done was to add a small piece of cardboard, THEN fiberglass the reinforcement.  The cardboard would provide a small gap between the rear clam and intake tube.

img_1216
So I cut the reinforcement out and started over again.
img_1218
This time I used a piece of foam core poster board to create a gap.
img_1219
I remember spending so much time on the first patch job – this took just a few minutes to do.  It’s amazing how versatile and easy it is to create and shape using fiberglass.  It’s pretty tolerant to a noobie such as myself.

Another item that needed a Go-Back was the recontouring I did of the rear wheel wells.  The initial recontouring just didn’t look good – it looked super amateurish.  So I threw the towel in and redid the entire wheel well (just about).  The new look is much nicer, but it’s at higher risk for tire rubs during suspension compression.  To test for rubs I removed the springs from my rear shocks and articulated the suspension until the shocks were fully bottomed.  To replicate this condition on the street would mean something went terribly wrong and I broke some stuff.

The good news is – the tire will travel throughout the entire suspension range and can rotate by hand at all positions.  The bad news is it’s close.  Really close.  In reality I think the motions of a wheel rotating at speed and a suspension and rear bodywork vibrating/moving will mean tire rubs are inevitable.  It’s difficult to determine how severe those rubs will be.  For now, I’ll leave it as it is but a v3.0 of the wheel well may be in the cards.  I really didn’t want to add fender flares but it’s the only way I’ll be able to get more clearance for the tire.

img_1228
New contour sanded back – it’s about 0.25″ thick at the edge.
img_1214
Suspension fully bottomed.
img_1215
Not sure a business card would fit in there!

OK – enough about things I messed up.  One of the pieces I’m most happy with is the air duct I made for the radiator discharge.  I haven’t quite figured out how I want the exit to look but know I didn’t like where I had left things.  Using some more poster board I made a simple set of “walls” to shape an exit for the duct.

img_1173
A simple discharge for the duct.  I used the poster board to construct straight, flat planes for the sides and added a slight contour to the fore and aft walls.
img_1174
Some peanut butter to patch and create a transition between the existing and new material.
img_1192
I used clay to help stabilize the foam board but it still wasn’t super stable.  Stippling and rolling the fiberglass to get all the air out was a bit difficult because the boards would move a little.  Not a huge issue as these walls aren’t structurally stressed but a little annoying.
img_1201
Poster board removed.  There’s just enough clearance around the edges so the front clam can lift straight up without interfering with the duct exit.  I decided to trim the discharge so it’s about 0.25-0.5″ below the surface of the hood.  This keeps the sides of the duct hidden and any fiberglass flaws should be harder to detect.
img_1206
Lots of trimming needed!
img_1207
The discharge is a lot bigger than it used to be!
img_1203
Another view.
img_1282
With the hood off and top trimmed.
img_1313
While I had the access, I cut out the opening for my super basic tow hook.
img_1335
Fit check with the front back on – the cardboard piece is my ghetto stand-in for a mesh cover.
img_1337
Almost looks like I had planned this whole black-stripe-running-down-the-center-of-the-car thing.
img_1320
The trick will be making the mesh cover so it has a slight contour to match the rest of the body.

Dumping radiator air just in front of the center of the windshield is about the worst place to do so.  The center of the windshield is a high pressure zone – which makes pushing hot air out of the radiator more difficult.  I plan to fabricate a small Gurney flap which I’ll be attaching to the rear edge of the hood (where the carbon piece turns down).  I’m hoping the Gurney flap will cause enough disturbance/mixing to lessen the high pressure zone immediately in front of the windshield, making it easier for the radiator air to flow out.

After a spell of feeling like I was aimlessly spinning my wheels I’m starting to feel like I’m making tracks again.  Getting the splitter laid out and completed and bonding in the front wheel well liners was a real step forward in closing out the work for the front end.  Neither the front nor rear liners are complete yet but I need to be careful with the last bits as they’ll really lock the clam geometries and stiffen them.  I need to get the aerocatches installed so I can get the preload right before bonding in the last pieces of wheel well liner or I might be fighting more twisted bodywork later.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s