Build "CJ3BL" 53 Willys

[CJ3BL]

Exhaust System

Overall specs: The headers have 1.5” primaries and 2” collectors. They v-band attach to the Y pipe. The Y is 2” into the Y with a 2.25” merge outlet, muffler, & tailpipe. All stainless for thermal and corrosion resistance benefits.

Here’s a bunch of photos that show how the headers and Y pipe turned out. I’ll then post a bunch of details on how they went together, and then follow up with posts on the back end exhaust routing.

These are some views of the headers installed on the engine. My plan was to have all the header primaries end up the same length, and I was able to accomplish that as well as I could measure/estimate the center line lengths. The driver side routing enables a one piece steering shaft. The passenger side snakes between the starter and engine mount area to the Y pipe run under the oil pan to the drivers side merge.



Here's the passenger side header before mounting it on the engine. The backside view of the flange shows how the tubes were fit to the flange port shape.
The tubes are welded on the inside of the flange for the full perimeter of the port and filed / lapped flat for sealing, plus there are two stitch welds to the flange on the outside of each tube for mechanical reinforcement.



Next are photos of the Y pipe. The installed photo shows the Y tube viewed from the front. In the foreground is the engine mount cross member, in the background is the bottom of the oil pan, and the Y tube sits in between. The Y tube has clearance of a little over 1/2" at the top edge vs the pan, and about 1" from the front of the tube to the cross member. I plan to make a skid plate that will mount at the bottom of the crossmember, then angle down to the bottom of the oil pan with clearance to the y tube in the angled section.

This was a learning experience!

 

[Jarod]

This is such a cool build! I really love the level of detail you are putting into it, along with the thought process behind it all.

 

[CJ3BL]

Thanks Jarod! I really appreciate the feedback. I sometimes wonder if I'm overdoing it with posting details. I sure learned a lot from other forum members' detail posts, so I'm trying to return the favor. Glad that you're finding it to be an OK balance!

 

[CJ3BL]

Y - Pipe:

This was (and is) the only stainless exhaust system build I've done. I tried to read up on it before I dug in.
It's all 304 stainless. The TIG wire is Harris 308 Stainless Steel .035". Flanges are from Stainless Headers MFG. Most tube bends from Cone, some from Burns. V-band clamps from Burns. Collectors from Cone. Flowmaster Muffler. Cone very good quality and less expensive. Burns tubing and collectors were more expensive, but their collectors are a work of art..! Their site has a lot of good info too. I erred on the less expensive side, since I was nervous I might just make a pile of stainless scrap. The combination of Cone and Burns products worked great.

Here are a couple of comments on methods used:

Argon back purge:
I did back purging of the interior of the stainless tubes with argon to prevent "sugar" oxidation in the interior (as before on the fuel tank). I picked up some rubber plugs at a local hardware store in sizes that would plug the ends for the back purge, and drilled one each with a hole for a piece of brass tubing I had around that fits the hose from a second argon tank, which is equipped with a flow regulator. A woodworking brad point drill worked nice for drilling the hole in the rubber, since the outer edges make a clean cut compared to a standard twist drill.
Here's some pics of:

• Drilling the rubber plug
• Flow regulator & argon tank
• Tube purge set-up

The plugs can work loose from the hose weight pulling on it, so I add spring clamps or some gaffers tape to support the hose when needed. When using the two rubber plugs, I’d leave a little gap at the solid plug to initially purge air out of the tube, then press it in and turn off the flow, to avoid wasting gas.
When needing to block off a tube for the back purge while welding near the end of the tube, aluminum foil was used instead of the rubber plug (which would burn). I just set a low flow and let the argon seep a little out the foil.

Cutting tubes:
I don’t have a bandsaw, so I used a cold saw (that’s what I call it, don’t know where I got that name from…) Specifically, it’s a DeWalt DW872 "Multi-Cutter Saw”. It's layout is like an abrasive cut-off saw, but it uses a carbide tipped 14" blade made for cutting metal. Instead of producing a bunch of abrasive grit it produces tiny metal chips, and doesn’t embed grit in the metal.

It worked great on the stainless tubing. The only challenge is that its fence and clamping capabilities are limited, so I drilled a hole in the deck and added a vice grip clamp and a little pipe clamp V attachment from a Strong Arm clamp. This gives some more options for clamping cuts on the curved round tubes, in conjunction with the fence as a back stop. Most of the time I could get the first cut right with the saw and just touch up the surface and then chamfer a little with a file.

Here’s a photo of the clamp arrangement on the saw:




Clamp for Welds:
For clamping tubes together for welds on the Y tube, I just used my vice to hold the two tubes together for tacking using the copper jaw covers, sometimes along with the same Strong Arm pipe V clamp attachment shown above. The vice and the pipe clamp attachment are also handy when tweaking the tubing back into round. Sections at the bends sometimes get a little oval after cutting and need to be tweaked back to round for a good fit. In the photo below, the left side of the tube was sticking out little, so squeezing it in the vice with the pipe clamp attachment supporting the right side allowed the left edge to push back in to round for a better fit up. The copper protectors shown on the vice work well to keep from marring the tubing, and for welding I attached the ground lead to the copper vice jaw tab rather than to the part. Very handy.



Y-Pipe:
I decided to layout and weld the Y-Pipe section first, as I thought the routing under the oil pan from the passenger side could be a challenge. It needed to snake around the starter, front driveshaft, oil pan, cross member, and steering shaft, and that route would heavily determine where best to position the collectors. Supporting it in place would be helpful in aiming the header routing too. As it turned out, I needed to rework the passenger side of the Y slightly as the headers were finalized, so trade-offs either way I suppose.

(I thought about buying those cool plastic header layout blocks to mock things up, but they’re pretty spendy, and would be used infrequently…if ever again, so I passed on that idea)

I started at the passenger side and worked towards the driver side Y merge area. This photo shows the Y tube run under the oil pan , behind the front cross member. It's supported by the jack at this point.


Here's a photo of the start of the Y junction on the drivers side. The two 2" tubes were flattened over a short part of the edge using a body hammer and dolly. They were then welded to each other to seal the inside of the Y. After this I added a short bead between the tubes running parallel to the tubes on each side.



The 2:1 collector was then welded to the two tubes above to form the complete Y. Welding this wasn't too bad except for the inside corners between the tubes- which were pretty tough to get the outer weld to connect fully to the interior weld joining the two tubes to complete the seal. With a lot of stick out on the electrode I got it figured out and was able to connect the inner and outer welds, so it turned out OK.
The next photo shows the assembled Y 2:1 merge section. The curved leg of the Y heads upward to a v-band clamp which attaches to the other curved tube shown on the right. That section will be the exit of the driver side header collector. On both the passenger and driver side, I used a short section of curved 2" tube to route from the header collectors through some tight spots to the V-band clamps attaching to the Y- pipe.



Wrapped up the Y-pipe. Here's some more photos with it in place.

Looking up from underneath, this is the passenger side, with the collector and output v-band in place to start on the header build. It wraps under the oil pan behind the front cross member. It's got 1/2" clearance to the oil pan at the top edge, and an inch to the cross member. Kind of tight in order to get as much clearance as I can to the driveshaft. The driveshaft u- joint ends up in front of the exhaust. With a 2" dummy tube in place for the drivehaft, the exhaust has about 3/4" clearance to the shaft worse case.



Here's a pic of the drivers side Y merge area, with v-band connection to the tacked driver header collector. The close camera position creates a bit of fish-eye effect - the photo looks curvier than it is in reality. The tube leading up from the Y merge to the collector V band is vertical. At the upper left is the passenger side tube coming from under the oil pan. There's a short curved section that brings the height down a little from the pan crossing level so the height of the drivers side is centered on the rail. This provides a little easier routing to the drivers side header.

After the Y merge, and the rear V band clamp, the exhaust tube then kicks up for a short distance over the transmission cross member. There's 1/2" clearance over the cross member and 3/4" to the floor. This area will need some floor insulation. From the engine to the xfer case area, the exhaust is hard mounted, then floats on rubber mounts the rest of the way back.

That's it for the Y, now on to the headers...

 

[1Sinner]

Excellent work as usual.

 

[CJ3BL]

Thanks 1Sinner! Much appreciated!

 

[CJ3BL]

Headers

Lots of tubes to cut and fit...



The headers have 1.5" primaries, with the thought being to prioritize low end torque. I found some nice stainless flanges from Stainless Headers Mfg. that have 1 5/8 rectangular ports that match the port shape on the 4.3 heads. To make that work, the entrance ends of the primaries need to be worked to transition from 1.5" round to the larger rectangle port shape. From what I'd read, folks usually use a purchased short stub section tube that's tapered to a larger cone end, and then form the rectangle shape by working it over a mandrel. The stub is welded to the flange, and then the round tube primaries are welded to the stub. Being a sucker for punishment, I thought I'd try a variation on this idea. Instead of purchasing and modifying short cone stub pieces and welding the primaries to the stubs, I slit the primary tube itself, spread and shape the tube to the rectangular port using a mandrel, then welded in a V shaped filler cut from the same tubing. This creates the flare in size and shape without thinning the wall of the tube. The V filler would be located on the underside so it’s not visible. This would look a little cleaner than having stub tube welds on every primary just beyond the flange. There could be some risk of a crack at the end of the slit cut, but I think good prep and weld will prevent this.

I made a mandrel from a piece of 1.5" square 1018 steel- first turning it on the lathe, then shaping the rest with a grinder and file:




Tried it on a test part. First a small hole was drilled to define the end of the slit and reduce the stress riser at the end of the slit. Then the slit was cut with a jig saw (with the blade cut short to avoid the tip banging into the opposite tube wall). The slit was then spread using a cold chisel as a wedge.The mandrel was inserted by hammering it in a little, then the tube sides were shaped with a body hammer. This was repeated until the full shape was formed to fit the inside of the flange. The photo below shows the formed tube inserted in the flange, before welding in the filler V. One shaped, the V filler section was cut from the same tubing, fitted, and welded to fill the bottom side of the tube. The V filler piece weld was filed smooth where it faces the flange opening, and little final fitting was done with further mandrel work. It worked pretty well, so I went with this approach rather than ordering a bunch of cone stub sections.



I also made some clamps, that are a cheap emulation of IC Engineworks clamps. I would have bought their clamps but they weren't available in 1.5” at that time. (They released 1.5” after I built the headers). The ones I made were less expensive and worked well enough. The clamps used muffler clamps and some un-plated steel hinges. Two clamps were positioned side by side, clamped to a straight piece of tube, to position them in line next to each other. The hinge was then welded to the base of the clamps.

The pics below show the clamp construction, in use, and the shim. They are probably clumsy and sloppy compared to the ICE clamps, but they work fine with some effort. Each side of the clamp attaches to its respective tube, then a welding spring clamp is used to squeeze the tube joint together. The main problem with this clamp is that the width of the clamp is pretty wide and the fit to the tube curvature can make some bend combinations hard to align. I think the narrow ICE clamp would work better as far as ease of achieving accurate alignment. Their less bulky design also allows better weld access than these clunkers. At the same time, these worked just fine with some finesse in adjusting their positioning - and they were cheap!



On to actually making the headers...

The drivers side was worked on first as it has to deal with the steering shaft and is closer to the frame rail too. The routing around the steering shaft was important to establishing a target tube length for all of the primaries. After a lot of playing with tube sections trying to envision shapes in my head, the routing plan took shape and cutting, fitting, and tacking began. While I considered block hugger style headers as they are more compact, I didn't go that way because I thought they would be hard to fit with the single, straight shot steering shaft planned, and that equal length tubes might perform a little better.

Here's some pics of the first two tubes routed and tacked to each other, but not tacked to the flange or collector yet.



More to come...

 

[CJ3BL]

Headers continued

Here are some more pics of the driver side header going together, with all of the primary tube sections tacked to each other, but not tacked at the flange or collector yet. The reason for not making the latter tacks was so that each tube could be removed and the tube joints finish welded individually while access is easy.



The tubes are 14" in length along their centerline, within about 1/4" as best as I can measure. There’s more than an inch clearance to the frame, and about 1/2" clearance to the steering shaft if I position the shaft at its original height at the floor. In the above pics , the shaft is just resting on the tubes since I it's not mounted at the firewall yet) There's room to move the shaft up on the firewall to increase header clearance, and I think I want to do that to change the steering wheel angle just a bit as well. Spark plug access is fine, socket fits easily with a u-joint on the wrench extension.

Looking ahead to the flange welds, some folks just weld the inside perimeter only, some weld the outside top and bottom edges in addition to the inside, and I've seen mild steel headers that were welded inside and brazed outside. I chose to weld the inside perimeter plus add a stitch weld at the top and bottom for mechanical reinforcement. The stitch weld segments leave the sides around the fasteners free of a weld bead so there’s no interference with the fastener or wrench fit.

With the tubes extending out around the steering shaft, they would hit the stock inner fender well. They are well away from the tires at full stuff & turning though, and I'm already planning on making new inner fender panels as part of the highline of the fenders. Will shape the rear inner fender panel to accommodate the headers. I may also include a cooling vent panel at the "triangle" behind the fender / in front of the firewall at the body side to help airflow over the headers and out the back of the engine compartment.

With the tubes tacked, I then pulled the header apart and finished the tube section welds on the bench.

After welding the primary tube sections, the tubes were final fitted to the flanges and tacked. At the other end, the first steps of the collector fab started with insertion of the primaries into a positioning tool that Cone Engineering sells. It's very handy:



The Cone Engineering collector can be built with an internal tapered "spike". It's fixtured for welding using a little tool they provide, that holds three sides of the collector spike, which is intended to improve flow at the tube transition inside the collector.



The sides of the spike get tacked, the tool is removed, then the spike is final welded. It is then positioned for welding at the junction of the primaries while they are held in position with the locating tool shown earlier. The spike seals the inside of the tube junction and also provides a smooth flow transition from the primaries into the collector. Some folks use a flat "star" piece here (with three corners in this case), which would be easier to weld, but I liked the spike design better as far as flow. Here the spike has been welded, and the collector assembly tool is removed.



With the spike welded in place, I ran a bead along the sides of the tubes at their intersection- which makes it easier to weld the indented area of the collector around the tubes. Then the collector was welded on to the primaries. The inside corner welds of the formed collector shell to the tubes were a little challenging. As in the Y merge, it took a lot of stick out and careful positioning to get the arc directed where it needed to be, but I was able to pull it off OK.

After the collector assembly came the final flange welds. I did the inside flange welds around the full perimeter of the port first. The tube ends were set slightly down inside the flange hole to form a v groove space at the flange hole edge to make sure there was good depth of weld to support subsequent finishing to bring the weld bead flush with the flange surface. After the inside port perimeter welds, the outside tube to flange welds were made top and bottom for additional mechanical support to take some load off of the inside sealing welds. Inside and outside flange welds are shown in the photos.

The photo of the outside weld also shows aluminum foil being used as a dam for the back purge of argon. The yellow purge hose has a brass tube clamped in the end, and that tube is clamped to the header flange with a clothespin style welding clamp, with the tube outlet pointing at the underside of the weld area. Then it's wrapped with Al foil to trap the argon, but leaving the joint area exposed on top for the weld. It doesn't need to be a tight seal - just enough to keep the back bathed in argon. This approach was used in several other spots, such as the backside of the collector spike weld.




Lastly, the inside welds were carefully filed to get them close to level with the flange. I then did the last little bit of leveling by lapping the inside of the flange on sticky back sandpaper stuck to my table saw top. Here's a pic of the inside flange surface after lapping:




More finished pics in next post.

 

[CJ3BL]

Header Wrap Up

Here are more photos of the finished driver side header:



With those done, moved on the the passenger side with the same approach. They were able to fit up with the same 14" primary length as the drivers side with all the tubes within about +/- 1/4" of that nominal length.

Below is an in-progress photo of the passenger side coming together. At this point the primary tube sections were all fit and welded, formed at the ports, tacked to the flange, and the collector spike welded in place. In the photo, the front tube looks close to the frame rail, but there's actually around an inch of clearance - it looks closer due to the angle of the photo.



At this point, I bumped into an issue. The collector fit the primaries great, but the extension from the formed collector piece to the y-pipe v-band flange (that I made early on along with the y-pipe) wasn't fitting as well as I would like. I stopped and didn't weld the collector.

I first tried cutting off the tacks on the collector part and making a new extension piece that was a little longer to improve the fit. This was an improvement, but still not the easy fit I was wanting - it took some force to line up the joint. To fix it, the mating part of the Y-Pipe needed to be revised as well. I wanted all of the v-band joints to slide together cleanly with no struggle as I didn't want the y-pipe joining the headers to place static stress on them, and also wanted the v-band mating flanges to be as co-planar as possible so they would seal well. I figured the headers and Y-pipe will see some stress from thermal expansion, so I wanted the static mechanical fit-up to be stress free, not a forced fit.

So, the previously welded end of the y-pipe was cut off and I made a new end piece with a new v band flange. To make sure I got it right, the headers were mounted with final gaskets and bolts torqued. I snuck up on the fit of the new piece trimming and filing a little at a time, checking alignment after each fit adjustment, then after each tack, and before final weld. The length of the span across the oil pan area makes any little misalignment at the weld joint on one side get amplified at the opposite side - so dialing in a really smooth fit was pretty fussy. In the end, all the final joints slipped together easily. The rework grief was worth it.

Here's pics of the finished passenger side header and y-pipe.



This photo is looking up from underneath the passenger side. I positioned the collector extension towards the engine mount to get it away from the starter to reduce heat transfer to the starter and to also tuck the Y-pipe close behind the cross member than runs under the oil pan. I may add a little shield for the starter. Note that in this shot it looks like the V band clamp hits the engine mount plate, but there's plenty of clearance - the clamp sits well below the engine mount plate.




Posted these before, but here's a couple pics of the finished installation as a wrap up:



All done with the challenging parts of the exhaust system!

 

[1Sinner]

Damn fine work

 

[CJ3BL]

Thanks once again 1Sinner! I really appreciate your feedback!

 

[CJ3BL]

Tailpipe, Battery, Center Section Sorting Out…

This post serves as a mini map for the next steps of the build. As described in an earlier post, the exhaust design, battery location, and “central frame stuff” like floor supports had interdependencies that were being worked out.

• On the exhaust, key decisions were to build headers, and run a Y-pipe at the engine to feed a single exhaust.
• For the battery (or batteries) , a key decision was to mount one (or two) in a protective box along the frame rail under the driver area.

These decisions meant that the exhaust system would route to the drivers side, but jog around the battery location. The battery box idea and access to the box also had implications for other stuff like seat mounts, center console, floor supports etc at the center of the frame - ie “center section stuff”.

With the headers and Y-pipe part of the exhaust now done, the back end of the exhaust, the battery mounting, and the other center section stuff interleaved a lot as the build progressed, and involved some sorting out - including some rework.

This graphic is an early view of how two batteries and the exhaust system would be located:



After a lot of valuable feedback and discussion with forum members, and thinking through the details I decided against this muffler arrangement. Putting the muffler so close to the batteries was too much of a heat concern. At that point, I was still thinking I might be able to fit two batteries as shown, but needed to move the muffler further back to reduce heating of the batteries. There was not a lot of room to move the muffler back given the rear axle and fuel tank location constraints. After some mock up work with the muffler I had already bought, I concluded that moving it back and angling it would work, so I proceeded with completing the exhaust system while I thought through the center section/battery mounting design.

Here’s a photo of the completed exhaust system based on that layout assumption, for comparison to the initial concept graphic above (sorry it's sideways vs the graphic)




With the above context, what's coming up will first be posts on the completion of the exhaust system. After that, the posts will address the battery / center section stuff, which had a bit of churn, but which eventually settled down for the better.

 

[CJ3BL]

Exhausted...

I had bought a FlowMaster "Super HushPower" muffler to work with when I was planning duals, and decided to continue with it. It’s a little long for the location I settled on, but I didn't want to change to a shorter muffler as the longer one helps keep the sound level down. It’s also supposed to have some internal insulation to reduce radiated heat, which helps the heat exposure concern for the batteries. To help with the fit, the muffler inlet and outlet were shortened a bit and then V bands were added for clamping.



Here's an in-progress shot of the central tailpipe section ready for tacking. The hinged muffler clamp tool is the same idea that I used on the headers, but in 2 1/4" size. The exhaust tubing is clamped to the rectangle tube section and angle steel pieces resting on the frame rails to establish the height of the exhaust relative to the rails. The lower edge of the exhaust tubing sits 1.75" above the bottom of the frame rails. The muffler is above the rails, springs, and driveshaft, so somewhat protected.



Here's some shots of the finished exhaust routing. Since the routing around the fuel tank, axle, and frame rail at the back was the hardest to fit, I started there, then joined up the central section last. Getting the axle area to fit well moved the muffler forward a little bit, but I think it will still work out OK as far as the battery proximity issue. (Later on I decided to just use one battery, so even better fit then)




Overall, I'm pretty happy with the routing. Here's some detail shots.
Under the T-Case & Overdrive:



Over the axle > turned to the side to clear the fuel tank > over the fully compressed spring > under the frame rail/cross-member > turned towards the back bumper… I fit this with the main spring compressed all the way to the frame rails, which inverts it as shown in the photo. In use, there will be a bump stop so there will be even more clearance above the axle.
From the side:



From above:




The tailpipe exits at the corner of the rear bumper The pipe has a short turn down to deflect the exhaust downward away from the bumper structure and the exhaust tip is cut parallel to the taper of the lower side of the rear bumper. The side rock rail extension feature of the rear bumper corner is outboard of the exhaust, providing some protection for the exhaust tip.



At this point, the exhaust was done except for the hangers. The hangers will get installed after building the battery box / crossmember assembly, since one or more hangers will attach to it.

 

[CJ3BL]

Center Section Stuff

With the exhaust system routing done, and the decision made to mount the battery next to the frame rail under the driver seat, it was time to make the battery box and other related “Center Section Stuff”.

The foundation for the center section assemblies is a crossmember structure that supports the battery box, the central floor, and other attachments. This foundation and the assemblies that attach to it went through two versions:

• Version 1 of the crossmember was removable from the frame- it bolted to 4 brackets. It initially was intended to mount two batteries, but early on changed to one. Packaging for two was cramped and complicated. The battery box was welded as an integral part of the removable cross-member assembly. Many other center section parts were also welded on as an integrated assembly - and it became too heavy and cumbersome to be reasonably removable…
• Version 2 of the crossmember itself is not removable - it’s welded to the frame rails. Elements such as the battery box bottom and top were broken out into smaller removable parts to re-partition the service-ability and weight. The change also included a number of simplifications and design improvements.

Here's a photo of Version 1 with attached assemblies, right before the rework effort. You can see the early rear frame "V" brace in the lower right corner of the pic (which was changed to an X brace as described in earlier posts). That change happened at about the same time as the center section rework. The removable cross-member / battery box stuff is beneath the seat mounting frames that mount to them. As you can see there is a lot of “center section stuff” just ahead of the old V brace.



Here's the Version 1 cross member / center section assembly unbolted from the frame. On the right is the lower battery box with skid plate. It had been welded in, but here is shown cut off in preparation for reuse in Version 2.



Without going into the Version 1 design details, here are the problems it posed that were all improved in Version 2:

• The removable cross-member and attached assemblies made it hard to seal the floor panels, especially at the attachment points to the frame.
• The assembly became too heavy and awkward to handle, practically defeating the removal feature.
• The removable design didn’t reinforce the frame as well as a welded cross-member.
• The battery wiring, center console electrical controls, and air system plumbing that would be mounted on the assembly would all need to be dismantled to remove the cross-member structure.
• The position of the forward edge of the cross-member prevented removal of the overdrive without removing the cross member.
• The Version 1 manual mechanical parking brake handle consumed considerable space. Version 2 replaced the big handle with a small switch - enabling limited console space to be devoted to other controls.
• With the electric e-brake, the mechanical cable and routing through the cross member attachment at the rear “V-brace” could be be eliminated.
• The Version 1 assembly formed the back of the x-fer case tunnel, but the mounting approach was clumsy with poor sealing.
• The Version 1 lower battery box with skid plate was welded in the assembly so was not easily repairable. In Version 2, it bolts to the crossmember, so it can be easily removed for repair or replacement when it gets too beat up.
• The battery cable routing over the top of the battery was awkward and forced a tall upper battery box, which then forced a tall seat position. The new design improves the cable routing, lowers the box height, and allows for lower seats if I choose to lower them in the future. (I’m happy with the height now, but time will tell)
• I wanted to add two parallel 200A circuit breakers in the battery cable at the box to protect against unleashing a "battery arc welder" if the cable gets cut or shorted somewhere. Version 1 had no space for this. Version 2 includes this and places the reset buttons within easy reach (but not easily visible). These also also serve as an anti-theft device.
• The Version 1 welded assembly was a little warped and not well aligned to the rear rail height for rear floor support. I could do better.
• The rear center seat mounts were cantilevered pretty far from the seat frames. Changing the design would provide a more direct mounting.
• I wanted to add capability for some mechanical shielding and/or driveshaft loop near the rear t-case output, to protect electrical and fuel lines routed along the rail in case the shaft flails around from a failed u-joint.

With the background for the Version 2 Center Section changes established, the next posts will describe the build. They are organized as follows:

• The Version 2 center section cross member structure & battery box will be addressed first. This leaps ahead in time relative to the original build sequence.
• Posts on the parts that interface with the Version 2 cross-member structure (like seat mounts and transmission/ transfer case tunnel) are covered in the original build order, but with later Version 2 rework included to bring the descriptions to current status.

Version 2 cross member & battery box build:

In the above photo of the removed Version 1 assembly, it shows that I cut off the bottom of that battery box to reuse on Version 2.
Here’s info on how the lower box was initially built for Version 1, plus its modification for version 2:

This photo shows the flat layout of the 0.120" 1018 CR side panels, ready to bend. (Parts of the top were later cut off in re-using it for version 2). The notches along the bottom are for water drainage at the skid plate.



Here's the formed box side panels, and the 0.188" skid plate, which has formed ramps at the ends to help it slide. The skid also has a formed lip to locate the battery on the side facing the frame rail. The box assembly will be positioned on the cross-member so that there is a 0.75" min gap between the battery and the frame rail. The space allows room to run the brake line along the rail, and also allows water to drain so it doesn't get trapped and rust out the rail.



Here it is welded up:



Re-use for Version 2:

Here are the details on the lower battery box modification for the Version 2 assembly. The lower battery box / skid plate was originally welded to the cross-member in Version 1, while in the Version 2 design, the lower box is removable and bolts to the underside of the new cross-member. The differences in the cross-member designs also enabled the old battery box to be cut off and still have enough material remaining to add a mounting flange at the new target height on the box.

Here's the reclaimed and trimmed to height box, with its new mounting flange:



Here’s the new flange welded on to the old lower box (box is upside down- skid plate on top):



More coming up in the next post...

 

[CJ3BL]

Center Section Cross-Member (continued)

Bosses were added to the bottom side of the passenger side tubes to mount a driveshaft shield or loop attachment to protect the wiring and plumbing that will run along the passenger rail should the rear driveshaft upper u-joint fail. After welding they were ground flush and tapped:



The cross member tubes, and lower battery box tube assembly were then fitted and welded, along with an upper battery box bulkhead plate. (assembly is upside down in the photo):



In the next photo you can see a passenger side plate added that mirrors the position of the battery box fixed bulkhead plate. A top plate was then made (sitting on the table to the left), which will be added across the top of the two side plates. These are all 3/16". They are welded to the cross member and also will weld to the anti-rock cross member tube. Double shear seat mount tabs will weld to the top plate for the center seat mounts. The seat mount tabs tie to the frame structure more directly than on the prior Version 1 design.

You can also see two holes to the left side of the battery box bulkhead. These are for two sealed protective bulkhead feedthroughs for the battery cables. The Version 1 assembly had the cables routed over the top of the battery. This Version 2 routes them to the rear of the box behind the battery then through the feedthrough fittings. This makes it easier to remove the battery, enables a lower battery box height, and provides cleaner routing in the box and to the shut off switches in the planned center console.



The side plates were chamfered for the outer top plate welds, and the welds will be ground flush to accept attachment of the upper battery box and passenger side cover. The top plate was also welded to the side plates on the inside as well - since two of the seat mounts weld to this plate and I want it solid! First photo is an outer side, after finishing. The second photo is an inside bead.



To seal and reinforce this center section. a 0.120" rear bulkhead was formed and fitted. The formed shape provides battery cable feedthrough clearance, and fits to the top plate centered under the seat mount tabs as a stiffener for the panel.



Here's the Version 2 cross-member assembly fitted to the frame, and with seat mount tabs added at the center.



The new cross member design enabled positioning of the center rear seat mount tabs closer to the seats, reducing the leverage on the seat frame mount. Here's a pic of the old seat frame mounts positioned next to the new frame cross member mounting tabs. The seat frames will be modified to fit the new center section mounting tabs, but that will be covered in posts coming up on the seat mounting overall.



A simple straight flange was welded along the top of the front tube for the back of the transfer case tunnel to attach to. Then the new assembly was tacked and final welded to the rails and to the anti-rock tube cross member. (The lower battery box is not attached in these photos as it was removed for weld access to the frame rail)



Next up is the upper battery box assembly…

 

[CJ3BL]

I forgot to include some info on the passenger side of the cross member in the last post...The passenger side of the Version 2 cross-member has a 0.75" rise for additional clearance over the driveshaft and exhaust. I wanted it to have a curved transition at this elevation change that the floor panel will also follow. Here's how it was made:

Cut and chamfered (jig saw, angle grinder, and files):



Formed on the press:



Trimmed and welded:



The rise keeps the outer end level with the frame rail top when installed, with the raise farther inboard to provide clearance over the exhaust and driveshaft. The level outer area keeps the floor level inboard up to the point it needs to rise for the clearance.

Now on to the upper battery box. Here's a pic of the upper box with lid. Build details in next posts:

 

[CJ3BL]

Upper Battery Box

Here’s a photo of the upper battery box without its lid. The assembly attaches to the cross member with some screws into the bosses in the cross-member tube that were made earlier. The flange on the lower side of the box provides a sealing surface to the floor panel.



The new box is wider and lower than the prior version to support better cable routing and mounting of circuit breakers. The sealed feedthrough fittings for the battery cables are positioned behind the battery (lower left in photo). This makes it easier to remove the battery, and I was able to drop the height of the box a little over an inch versus the Version 1 box, so the seats could be set lower if I choose to change their position in the future. The battery kill switch will be at the center console and the new cable routing works nicer for that wiring.

At the outboard side of the box, you can see the back of the two 200A circuit breakers that will be wired in parallel on the hot lead. The wider box provides lots of room for wiring these. They are made by Buss, and the panel mount design is nice as the cables attach inside the box, while the reset actuators are on the outside of the box- easily reached under the drivers seat.

The whole thing vents/drains along the frame rail through the lower box/skid plate. I plan to use seam sealer and possibly a sprayed on coating or undercoat inside the box to seal the stitch welds and to reduce potential for corrosion.

Here's some step by step details:

The base plate is made from 12g (0.105") material. It has a bend that creates a bulkhead panel at the back of the box. I formed it and drilled the holes to mate to the threaded bosses in the back part of the cross member, then welded and finished the front/side flange section as a separate piece. More mounting holes were then added to attach to the front part of the cross member, and along the side to attach to the floor panel.



The other three vertical sides of the box are formed as one piece in 14g (0.075"), with 1/2" radius bends for rounded corners. I went a little heavy on the panel thickness here to make it easy to stitch weld to the upper and lower 12g flanges, plus to have a reasonably rigid panel to mount the CB's to on the outside, and to bolt to the cross member bulkhead flange on the inside face.



As shown, the back corners of the side panels were formed only at the top- so the lower part of the side panels continues rearward. The outer panel joins to the rear bulkhead panel, and covers up the anti-rock. It will have a flange welded on to seal to the wheel well tub- once I make the tubs. The inner side panel also continues back to attach to the rear bulkhead formed in the base plate.

The next piece is a 14g top plate that finishes off the upper opening of the box and which is co-planar with the center section top plate. The front of the rear floor will attach to this area. Careful fitting was need to make sure that the height matched the center section and rear frame rail tops - it's a 5" floor rise height from front to back like on a stock Willys.



Here's a shot of how the rear corners and top plate go together. The top edges of the sides were trimmed (as marked) after the welds were done.



Then this was fit and stitch welded to the base flange/rear bulkhead part:



The above also shows the holes drilled and finished for the circuit breaker mounting. The inside panel of the box was then drilled to match 1/4-20 threaded mounting holes on the cross member upright bulkhead panel, and a clearance cut was made for the battery cable feedthroughs. The feedthroughs attach and seal to the crossmember center section, so they don't need to seal to the box panel itself. The cut-out allows the battery box to be removed for clean-up or re-painting without having to remove the cables and feedthroughs.

Here's a detail of the panel cut out around the feedthrough parts:



Last was to fit and stitch weld an upper flange, also from 12g. It has 1/4-20 SS nutserts in the corners for mounting the cover lid. I'll put foam weatherstrip around the edges of the flange, inboard from the lid mounting screws, to seal the box. The lid is made from 18g (0.048"). I made the top flanges wide, so that if additional hold down screws are needed, it will be easy to add more nutserts to the flange.



Here's the upper box in place with the battery and hold down strap. The hold down strap is a simple formed flat strap that slides over the battery and has two screws into the cross member bosses to hold it down.



Next up is fab of the lid.

 

[CJ3BL]

Battery Box Top Cover

Here’s how the box cover turned out:



Some details on making it:

The lip around the edges is hammer formed. Usually folks step roll or bead roll the panel then do any forming of the edges - so the formed edges don't interfere with the throat of the bead roller. These edges are so shallow though that it worked fine to form them first, and I liked doing them first as I could better visualize what I wanted to do to the top of the panel, and not having the step formed made clamping the hammer form caul easier.

I didn't make a full perimeter hammer form,. Instead, I just made one corner from a misc piece of steel I had around. It has long sides for forming the edges but is still shorter than the overall shape so it would fit inside the cover as it was formed. The caul is aluminum, and has one corner radiused to support the panel while forming. The caul piece is bigger than the panel, but the size doesn't matter since the lip is formed away from the caul. I moved the panel around on the form to make the full edge on the panel. The form edges are rounded a bit for a slight radius to the bend, and the box corner radius is on the form and caul corners at the lower right in the pic:



I completed the basic edge forming, then rolled a simple step form on the top of the panel using a step die on a bead roller.

I was debating between running beads versus a step rolled design. The Version 1 battery box had straight flat top style beads. but I decided against that idea this time because a nice bead layout on the box cover would not align with the bead pattern that I’m thinking I’ll make on the rear floor - together they’d look misaligned (even though mostly covered by the seat). I also tend to think of the flat crowned bead shape (like the original Willys rear floor beads) as being appropriate for load bearing panels like a floor that has stuff being loaded on it. No one's going to be standing on the battery box cover. That got me thinking about doing a simple step design instead.

I had made straight steps to join panels before, but not a decorative step with curves, so I played around on some scrap to try some ideas. Using standard step rollers I was not having very consistent luck staying on course around curves. Gojeep on the old forum hipped me to the idea of using a tipping roller on top, and sure enough it made it easier to follow curves. Here’s a link to a mailbox project he did that shows how he did his much more challenging rolling project: Custom metal bead work practise. Thanks Gojeep!

With that insight, I settled on using the standard step roller on bottom, and a tipping roller on top- set up so the step pushed downward (with formed panel edges up)- which let me follow the rear tipping roller edge around the pattern marking. This was easier to see and worked great.
I briefly entertained the idea of deploying what I learned on a fancy design like a step rolled willys W, but in the end decided to keep the design really simple - so I didn't screw it up, and because it fit the uncluttered look I'm striving for. Here's a pic of the roller die set-up with the freshly rolled part, and a view of the part from the top:



With the top design rolled, I then finished cleaning up the formed edge. The initial forming and rolling left a little convex dip / oilcan along the sides of the panel surface because the lip on the panel stretched a little. You can see this in the top view photo above, where the right rear corner is lifting a bit. I used a shrinker to very lightly shrink the formed edge to return the panel shape to flat:



Then worked the corners and edges a little more on the hammer form, planishing the shape for consistency and finishing with a fine file and light sandpaper. Fit the cover on the box, and drilled the mounting holes for the 1/4-20 button head hold down screws to wrap it up:



Some details of the floor panel under the box need to be revised, but I think I’ll be making new floor panels for a variety of reasons. Next up is making a panel that fits to the cross member in the same area on the passenger side.

 

[CJ3BL]

Center Section Passenger Side Panel

The next step was to fab this passenger side bulkhead / riser panel, that attaches to the cross member. It's essentially the floor in the step up area on the passenger side.



Here it is installed (left of center), along with the center section floor panel:



Before fab of the bulkhead panel, I also made the center section floor panel shown above, as the bulkhead panel needed to be fit to the top of the floor panel.

The floor panel extends the front passenger floor over into the center section. This new piece has some bits that patch the passenger floor panel where changes had occurred, plus it adds flooring to the center section. I needed the sheet metal thickness in place at the patch areas before building the bulkhead/step panel. The main passenger floor panel is pretty cobbled at this point, and I'll probably make a new one once all the changes are worked out, but for now decided to include patches in this center panel. (I'll likely keep the new center floor panel =if I replace the main passenger floor - just trim the patch bits off since they'd be incorporated in the new main panel).

The center floor has two bends that allow it to closely follow the rise in the crossmember over the driveshaft. The shape is a sharp radius bend up at an angle , followed closely by a wider radius bend to place the floor surfaces parallel with 3/4" difference in height. Had to puzzle on that a bit to come up with a bend approach, and got lucky...

I bent the sharper bend angle on my little cheap angle iron sheet metal brake. To get the second, wider radius bend I used a radius bend form that I made for that brake. It's made from a section of 1.75" diameter tube, with a square tube welded along the back to provide some rigidity. However, the position of the bend is too close to the first sharp bend to be able to put the part in the brake. Hmmm.... I ended up clamping the first sharp bend of the floor into the recess along the top of the large radius bend form using a piece of bar stock and clamping the whole assembly to the top of my table saw bench. Then used a chunk of angle steel along the panel to bend the large radius by hand:



This worked like a champ, and the panel fit the crossmember profile nicely. Here it is fitted in place- the bend area is in the middle of the pic. (I'll weld the panel to the frame rail top and cross member upon final assembly). At the back edge there are three 1/4-20 threaded holes to attach the bulkhead panel flange to the floor. The two inboard ones align to nutserts set in tabs welded to the cross member. The outer hole aligns to a welded and tapped blind boss in the 2"x 3" outrigger tube that runs between the frame rail and rock slider.



With the center floor in place, the bulkhead panel that attaches on top of the floor could be fabbed. It has a 5" rise height to match the center section top panel seat mount area. The front edge of the rear floor panel will align to that surface. The combined width across the battery box, seat mount plate, and passenger bulkhead panel also need to closely match the rear floor width of 36". The flange along the bottom of the passenger bulkhead panel needed to have a 3/4" jog to follow the underlying jog of the floor height. The floor panel is 0.048". The bulkhead material is 0.075" as I wanted it to be pretty rigid to support the rear floor.

The 3/4" offset of the flange bends and the thicker material posed some additional challenges for my meager metal working skills and toolset to puzzle over... I wanted to form the two straight but offset flange bends.

The simple angle iron brake I have has one long bar- so it can't handle the two bends with only a 3/4' offset. It's also too wimpy for the thick material. The SWAG finger brake set up on the HF shop press can bend the thicker material easily, and handle separate short bends, but it didn't have enough clearance to the base to enable the first bend without damaging the offset second bend area of the panel. I decided to make a gadget that would give enough clearance to make the first bend without damage to the second bend area. Here's what it looks like:



It's made from a U section cut from of 1/4" wall 2" x 6" tube. The bottom of the U sits against the lower angle section of the finger brake fixture. Two plates are welded to the ends of the U to stabilize its position in the lower angle. Lastly, there's a pice of angle stitched on top. Kind of a kluge, but it worked. The added height provided enough clearance to the base to allow the panel to dip down with no interference while forming the first bend- avoiding damage to the area of the second bend.

Here's a pic showing the second bend being formed after the first one was completed:



Once the bends were formed to make the 3/4' offset flange sections , a little filler piece was made to join them at the floor rise. It was bent to follow the floor shape using the same bend method as the floor. Here the filler piece, fit to the floor panel and marked to trim for welding to the bulkhead panel flange:



Then end plates were fitted and welded, and holes drilled for mounting screws. The hole positioning was pretty important to positioning of the panel overall, and I wanted the hole fit to be pretty tight tolerance. I turned a screw tip into a point on the lathe and used it to mark the hole positions. It was threaded in the threaded mounting holes from the opposite side of the panel and then the panel face was tapped with a hammer to mark each hole with the panel clamped in place. On the one blind threaded boss, I used a set screw with a similar point. The lower flange holes were drilled first so the panel could be screwed down tight to the floor before marking of the end plate holes. Here's a pic of the marked end plate with the pointed tip screw used to make the marks:




Here's the fitted and trimmed panel. The outer plate will have a curved flange added to the back edge once the wheel tubs are built. the flange will seal to the front edge of the tub.



I was glad to have this part done…

 

[csutton7]

I'll say it again, awesome workmanship!!

 

[1Sinner]

Yup. Straight clean. By comparison, I'm a hack.

 

[SHARPMACHINE]

Hol D Cow, she's a beaut!

 

[CJ3BL]

Thanks a ton csutton7, 1Sinner, and SHARPMACHINE! Don't knock your own work 1sinner! Your rig is killer, and I've learned a lot from your build posts!

I'm trying to crank through recreating my build thread to post it here on Irate. While it's taking some work it's been kind of fun to refresh my memory on the decisions made along the way. Lots more posts to come... then the pace will slow to a crawl to match my actual work speed, which is pretty dang slow given my tendency to overthink stuff.

I really appreciate your comments! They are very motivating to keep it moving forward!

 

[CJ3BL]

Center Console

Version 2 of the center section stuff includes a new rear control console section. Like version 1, it has an overall battery kill switch, a separate battery kill switch for the winch, and two pneumatic valves to control the ARB's. Getting rid of Version 1's mechanical parking brake lever frees up more room for the electrical switch panel that will mount forward of this piece. It will attach at the flange that's forward of the ARB switches. Here's a pic of the completed rear console section.



Here's some details on how it went together and what's going on in the layout.

The console panel is 12 gauge (about 0.100"), as I wanted it pretty rigid. The bend at the top back edge has a 1/2" radius to make a smoother surface for handling. I bent the large radius first, then bent the mounting flanges at the base. That made it easier to get the height and slope set right to match the seat frames (4 degrees). The little fixture I had just made for the shop press came in handy as it enabled the back flange bend after forming the radius corner bend- the part just fit, hanging down the back of the press:



After forming the first two bends, I cut and finished the holes for the controls, and then formed the front flange. Here's the part at that point:



Then fit side plates to support and stiffen the top plate. The power switch holes are pretty big, and the top plate would be prone to bending without additional support. Here they are chamfered for stitch welds and tacked:



Welded the sides, then made the front support plate, shown in the next pics. The little "bumps" at the front corners next to the top flange are there to avoid having a gap under the corners of the front switch console that will mount over the front flange of this assembly. That will have 1/2" radius corners so there's no sharp corners to rub on the seat fabric or be rough to the touch when reaching for switches. The positioning of the pneumatic switches and the front support were set based on fitting the plumbing underneath, providing good hand room behind the raised front console section, and to leave room in front of the support panel for the front switch panel wiring bundle routing, and possibly hang some other electrical parts on the support panel itself if needed. I didn't add any additional stiffeners to this panel as I wanted to minimize protruding surfaces that could rub against wiring and plumbing.

The first photo shows the fit up, in the second it's clamped in place for welds:



The finished part mounts with button head screws into 1/4" PEM nuts in the center section cross member top plate and floor panel.

The transmission / transfer case tunnel will attach to the flange at the front of the cross-member (at the lower right of last photo). The tunnel build will be covered in upcoming posts, including the modifications that were needed to fit this version 2 center section.

The open space between the tunnel attachment flange and the front panel of this rear console will be used to route wiring from the front switch console (which isn't built yet). I plan to make enough of a service loop in the wiring routing that the front switch panel can be removed without detaching its wiring, so the tunnel can then be removed for access to the full drive train from the top.



Here's a few comments on the rear console controls:
The battery kill switches are Blue Sea switches - these big ones are rated at 600A. They are made for marine use with a sealed package that can mount either in front or behind the panel. They make a little smaller, lower rated version too, and I’ve seen those on wheelers, but I wanted the higher rated ones for some margin. I’ve heard they hold up well.

Pigge on the old site Willys section had a similar pneumatic control set up which inspired this layout. I liked the idea of direct mechanical valves for the locker controls. Meiser on the old site tipped me to the Humphrey pneumatic valves following my inquiry with Pigge about the ones he used. Both types are a nice fit to the application. I went with the Humphreys as I like their small size, control lever shape, and plumbing routing.

As you can see in the photos, the air system was also taking shape in the space under the passenger seat, so plumbing the mechanical valve controls in the center console should be pretty clean.

 

[CJ3BL]

At this point, the main Version 2 center section change was complete. The rework of attached assemblies to fit the Version 2 center section will be covered in the build posts on those assemblies, primarily the seat mounting frames and transmission / transfer case tunnel.

A lot of other stuff got done prior to building those assemblies though, so I'll turn the time machine off now and the posts will return to chronological order.
When the exhaust system was done, the next thing that was worked on was fab of a new firewall, rather than patching and modifying the remnants of the old one - so the next posts will be on the firewall build.

 

[CJ3BL]

Firewall

The old firewall was pretty swiss cheesed. I also wanted to change it to extend farther downward to provide more legroom and space for the brake and clutch pedal throw in the footwell. The roll cage will have some forward extensions that head through the firewall into the engine compartment. The rig will also have firewall mounted pedals. I decided to build a new firewall with heavier than stock material - 11 g steel for the outer panels, and 14 g steel for the center recess. Part of the thinking is that the thicker outer panels would negate the need for the typical reinforcement plate used on a stock firewall at the pedal assembly.

For the center recess, I wanted to mimic the factory look with a broad radius at the back and a tighter radius at the front edge transition. I decided to simplify the top of the recess though - the compound curve at the top of the factory firewall is really nice, but challenging to make with the thicker material chosen. So the top edge design was simplified.

Here's the form I made for bending the broad back curve, and pics of the forming. It's a heavy wall 3" dia. tube with a few bits of 3/4" tube welded on top to locate it in the press. (It will also be used for the back corners of the body - which will have a smaller radius than the factory body).



Angle iron, clamps, and hammer were used to bend the tighter front bend that transitions to the outer panels. After bending, I cut off the excess to leave just the radiused edge to transition to the outer panels:


This edge was then chamfered fully for welding to the thicker outer panels. I aligned the panel edges at their inner surfaces for welding, with the thought being that the outer weld bead would help transition the difference in panel thickness reasonably by blending with the formed radius. Here's the first side panel tacked ( the right side panel needed some further adjustment of the fit before welding), and also the fully welded panels:

 

[CJ3BL]

Firewall continued + cowl

I had been keeping the cowl / firewall pieces of the old body together for body/chassis planning, but now that I had the plan and measurements figured out it was time to fit the old cowl and new firewall together. The front edge flange of the cowl that joins to the original firewall had some wavy forming from the factory, some funky spot welds that had failed, plus some big gaps at the firewall joint. I planned to clean up the edge to improve the fit-up, then final trim the firewall to fit the cowl. I also planned a perimeter flange on the firewall similar to the stamped original except it would be made from 1/8" flat stock welded to the edge of the firewall panels. It would have tapped holes or PEM nuts to screw the cowl flange to the firewall flange. This will allow the cowl to be removable so all the wiring behind the dash can be easily accessed from above.

I wavered a bit between this removable cowl idea and the more usual idea of making the dash removable. The removable dash is great but is kind of awkward to hinge down or just leave hanging when you work on it, and needs more of a service loop in the wiring. I think I can make the removable cowl work, and it seems like it would be nice to be able to wire everything fixed in position, then put the lid on. A downside is that with a top on the rig, I'd have to remove the soft or hard top, fiip the windshield down, and then pull the cowl to do major changes using the access. I’d also have more to mess with to seal it.

One thing I was sure of was that I didn’t want to be twisting myself into a pretzel wiring it from underneath again - especially with the raised transmission hogging so much space where I’d need to lay! As the work progressed I committed to the removable cowl option and crossed my fingers that I could make it work.

Here's a pic of the cowl. To remove it, I cut the firewall side of the joint so I could grind off the spot welds from the firewall side to preserve the cowl flange. I'm about half way done in this shot.



I found it worked pretty easy on the spot weld removal to hit the spot weld with an angle grinder. As it gets close to breaking through, the area surrounding the spot weld tends to heat and turn blue before the spot weld does- since the other panel it's attached too acts as a heat sink at the weld. Once it gets to that point, it was easy to peel the panel away as the surrounding area breaks free leaving the spot weld only. Then it can be taken down with the grinder a little further to be flush with the surface. I like doing it this way vs a spot weld cutter as it leaves the metal fully intact. It takes a while though… On the heavier brackets on the ends I used a spot weld cutter as the grinding technique doesn't work as well on the thicker material. This left the cowl flange surface in pretty good shape to work with a hammer and dolly to clean it up for fitting.

Here's a shot of some spot welds after panel removal, and another with the cross cutters I used to grab and peel the bit of panel back once the spot area was thinned with the grinder. They worked well as rocking them to the side after grabbing the scrap piece uses the shape of the cutting head like a miniature pry bar.



Pedal considerations:
I also started thinking through the pedal assemblies during the firewall work. My original thought on the brake and clutch pedal assembly was to see if I could use two original frame mounted brake pedal arms positioned as hanging pedals, and use original pedals with the shafts shortened. However, the arm ratio/angle and fit wasn’t working out, so I dropped that idea.

For the gas pedal I had bought a Lokar pedal assembly used on street rods. That would change later...

The gas, brake, and clutch pedal space is tight given the transmission is mounted high for a flat belly. That ,and lowering the floor angle bend location on the firewall as I was doing further compounded that situation - especially when using a manual transmission as the clutch fork was competing for space, as was the corner of the bell housing.

I planned to move the steering column outboard about an inch (and had included that assumption when mocking up the steering shaft location while building the headers). The body tub is 1” wider on each side, so the seats can move a little too - but the body width at the footwell is still the stock width and the transmission now takes up more space - so the pedal room is tight.

This pedal topic mulled in the background while firewall work proceeded …

For the top edge of the firewall, a 3/4" wide x 1/8" flange was fabbed, similar to the formed lip of the stamped original. I took my time in forming the flange to get a tight fit to the profile of the cowl, including angling the sides like the original firewall flange. Once this was fit to the cowl panel, then I clamped the cowl panel and firewall flange together and used them to mark the edge profile of the firewall for trimming. After trimming and fitting, the flange was welded to the firewall panels.

The following photos show the cowl and firewall flange clamped together and to the firewall panels - ready for initial tack welds. The clamping carefully positioned the inside edge of the firewall flange so that it barely overlapped the firewall panel edges. This created a nice V groove between the edge of the firewall panels and the flange to assure good penetration of the welds, which would be made from the inside of the firewall. With the parts clamped as shown, a couple of tacks were made at the end corners and at the center inset area to set the position of the flange to the firewall before flipping it over for final welds.


The cowl was then removed, and the lightly tacked assembly was flipped over to access the inside firewall panel to flange joint. With the weld side up, and checking the fit of the alignment before each tack, the flange was tacked to the firewall panels at 1 inch intervals. Once fully tacked, the final welds were completed in 1" intervals working from the center outward, alternating sides to minimize warpage.



To close the upper part of the firewall inset panel, the top of the panel was rounded so that it would flow into the joint with the flange at the top. A simple forming buck was made from a piece of 2x4, and shaped into the curved shape with a hand plane. The radius of the curve was chosen so that the existing panel material would largely complete the bend. This would enable a flat or lightly shaped filler piece to then be fitted to close the remaining hole.

I made two estimated pie cuts at each corner curve of the center panel, erring on the narrow side. The firewall was then clamped to the buck with a piece of tube on the panel to keep it from deforming while the curve was shaped. As the curve was formed towards the buck using a leather faced panel slapper, the pie cut edges started to overlap a little as seen in this first photo.



When they reached this point (without going too far and denting at the overlap) then they were opened a little by just running a jig saw through the slight overlap to open it up by the width of the blade (removing the buck to make the cut). It took two rounds of forming and pie cut adjustment to get the final fit. This panel is pretty heavy stock so in addition to the slapper, it took a bit of body hammer work to detail the compound corners. Once the fit was good, the pie cuts were welded up, and followed by a little clean up trimming of the remaining opening at the corners.

Here's two photos of the formed and welded center panel:


Filling in the top next... Overall, I like it. I'm trying to make it look a bit like the original, but with the inset area more centered around the engine and with a lower transition of the footwell angle bend to provide more depth of foot space behind the pedals. So far so good.

 

[CJ3BL]

Firewall continued

A simple panel was made to fill the remaining hole at the top of the firewall center recess, which was then tacked and welded in. Then smoothed the welds. Here’s some photos of it at this stage. Not all the way done on finishing of the recess welds, but it’s ok for now. I haven’t smoothed the welds that join the center and side panels at this point. I haven’t decided whether to leave them as is or smooth them out. I’ll decide before paint…



Next up, I needed to figure out how high to mount the firewall relative to the frame so that the cowl and hood are in the right position relative the the grill. This was needed to then determine the position and shape of the firewall opening at the bell housing / transmission tunnel, determine the position of flanges for fender mounting, and to design the supports that will mount the firewall to the frame.

I set-up the hood with the cowl bolted to it, resting on the grill with grill webbing at the front, and then supported it with wood blocks and shims on top of the engine to set the hood level front to back, and the cowl level side to side. Once I had it mocked up and level, I measured from the top of the frame rails to the bottom of the cowl flange that mates to the firewall, at the center of the cowl. This height was then also the reference height for positioning the top of the firewall at the firewall center.

With the firewall height determined, the height and shape of the bell housing/transmission opening in the firewall was set. I initially formed a flange for transmission tunnel attachment by hammer forming the edge on a wood buck.

The flange shape and tunnel fit turned out to be troublemakers. The root issue was really the packaging challenge of pushing the drivetrain up for a flat belly, with a manual transmission clutch fork, and with the narrow footwell of a flat fender. I asked for the trouble! Got it worked out though.

Here’s a brief outline of how this issue progressed, and where it ended up as far as the firewall build.

• The flange forming itself went well. The first photo below shows the formed flange. The clearances at the bell housing were set to be sufficient for engine movement, but intentionally small to maximize foot space. After checking fit, I decided it was too close, as access to some bell housing bolts was too difficult. (Second photo)
• Re-formed the flange to increase clearance and access. (Third photo). I was happy with this initially. However, when I later built the transmission tunnel, getting a clean fit of the tunnel to the formed flange was difficult.
• The clutch fork and bell housing clearance / bolt access were key constraints on the firewall flange and tunnel fab. In the end, I decided to change to a Tilton internal clutch throw out to remove the clutch fork constraint. That enabled a better tunnel design and fit to the firewall (although introduced a more complicated clutch set up). To support that final tunnel design, I decided it would be better to cut off the formed firewall flange and weld on a new flange piece. Hammer forming the firewall flange a third time would have made a mess of the panel. I’ll cover the weld flange version in posts on the tunnel build. (The last photo here shows a preview of that tunnel).

With the firewall mocked up in position again on the frame, the height was marked for where the cowl mounting flange transitions to some fender /vent mounting flanges. The fender / vent mounting flange and firewall supports to the frame are different than the stock arrangement:

• The firewall supports will be at the outside edges and weld to the frame rock slider tubes. The triangular support structures of the stock body that bolt to the frame rails are eliminated. This opens up access around the engine/bell area.
• The stock fenders bolt to the angled front flange of the body side panel, which extends forward of the firewall and inward into the engine bay. I’ll be replacing the forward / upper triangle portion of the body panel with a vent panel that bridges between the firewall and the back of the front fender.

The firewall frame supports will have the same fender mounting slope as the stock body panel, but that angle slope will not extend as far forward as on the stock body. Instead the support slope will transition to a vertical flange joint at the front face of the firewall. The body panel that will attach to it will have the same shape.

The vent panel surface will be parallel to the frame rails- so it doesn’t encroach inward into the engine bay as much as the stock body. This leaves more room near the brake and clutch cylinders, provides better engine bay access when the fenders are removed, and vents heat out the back sides of the engine compartment. The vent idea is similar to some aftermarket fenders on TJs. I’m still deciding whether the vent will be a separate piece, or be a welded part of the fender. In either case, it can be removed with the fender for better than stock access to into the engine bay. I may also use the vent for a fresh air intake duct on the passenger side, but haven’t fully decided yet.

Here's pics of the lower fender/vent mounting flanges welded in:



The lower section of the flanges will be trimmed to height and welded to the lower supports that attach the firewall to the outrigger rock rails.

Next up are those firewall supports...

 

[CJ3BL]

Firewall continued

Mounting the firewall to the frame structure was next. In addition to supporting the firewall, the supports will also provide support of the angled footwell part of the front floor. I rigged up some spacers and clamps to hold the firewall assembly at the right height and vertical position, then worked up the corner supports to fit. There’s a lot of angles to set in making the support: the fender slope angle on the front, the footwell floor slope at the back, the body taper angle on the sides, plus the support needs to work with the way I plan to mount the body side panels. I didn't think I could dial in bends accurately enough to form the support in one piece, so I made a welded structure from flat pieces so I could dial in the angles more easily. I first tacked the front and side plates in place while the firewall was clamped to the frame to get the positions right - then removed it for easy welding:



Then with the assembly clamped back on the frame, the partial support assembly was welded to the frame rock slider tube - (before final fitting of the rear plates):



You can see in the last pics that there's a gap between the top of the frame slider rail and the bottom of the firewall support side plate. The side body panel will have a flange formed along the bottom to attach it to the top of the rock slider tube. The gap in the firewall corner support allows the body panel flange to seal along the full length of the rail including at the corner support.

With the front plate of the support fully welded to the frame slider rail, the rear plate was then fit and welded. The completed assembly is very sturdy even with the gap along the side for body panel mounting. The top of the firewall will also be supported with a cage tube at the corners - tied in to the engine bay cage and the A pillar.

Here's the finished corner:



The side plate position on the 2" x 3" side rail is set so that the body panel will mount 5/8" inboard of the outer rail edge, with the mounting flange facing inward (into the gap of the corner support). The 5/8" protrusion of the rock rail beyond the body panels will continue all the way back to the rear bumper.

Here's the attached firewall assembly. There's a few detailing items remaining to take care of once I make floor panels, transmission tunnel, and mount the cowl.

 

[1Sinner]

You have way more patience than I do. It looks fantastic.