Through the dash cage builds

[NashvilleTJ]

reviving this thread.

I'm starting to plan for the through-dash tube and I'm open to idea's on how to best tackle, getting my angles correct..

1. It looks like the tube originating from the floor and coming through the dash would be 100% plumb, or parallel to the TJ dash/perpendicular to the floor.
2. Initial bend could be at the top of the dash and mimic the angle of the windshield frame.
3. A second bend might be required to align the tube to meet the header bar, before notching.

It seems kind of difficult getting both bends & the angle of the notch in the upper tube dead nuts without some expensive trial and error. Is the upper "notch" best attacked by leaving the tube long with multiple test-fits after some careful die-grinding? Thoughts on a round tube clamp at the header bar so the entire cage can be removed without having to tear apart the dash?

I cheated a bit. I tried to do that tube in one piece with two bends. After I got the windshield bend right, I then tried to do the top bend - and was not happy with it. Here is the initial attempt:

I was not happy with the alignment with the windshield, but the bigger problem was that the upper bend interfered with the door surround mount. I wound up punting and doing this in two pieces - sleeving it. Here is the result:

This would certainly not pass a cage inspection for a race vehicle, but it did save me a lot of trial and error (and material) had I continued to try to hit the mark with one piece and two bends. Both bends are tough, as they are both offset in two directions.

If Blaine ever pokes his head in here I'm screwed, but it worked and was much easier doing it this way.

Edit: Oops - I see Blaine posted while I was posting this.

 

[mrblaine]

If Blaine ever pokes his head in here I'm screwed, but it worked and was much easier doing it this way.

Bit late for that.

 

[sierradmax]

This isn't terrible.
View attachment 684408

Very clean, simple, but I'm not sure my skillset is up to that level without thousands of dollars of tube bent under my belt.

 

[sierradmax]

I cheated a bit. I tried to do that tube in one piece with two bends. After I got the windshield bend right, I then tried to do the top bend - and was not happy with it. Here is the initial attempt:

View attachment 684406

I was not happy with the alignment with the windshield, but the bigger problem was that the upper bend interfered with the door surround mount. I wound up punting and doing this in two pieces - sleeving it. Here is the result:

View attachment 684407

This would certainly not pass a cage inspection for a race vehicle, but it did save me a lot of trial and error (and material) had I continued to try to hit the mark with one piece and two bends. Both bends are tough, as they are both offset in two directions.

If Blaine ever pokes his head in here I'm screwed, but it worked and was much easier doing it this way.

Edit: Oops - I see Blaine posted while I was posting this.

I did not notice the sleeve to your photo posted 07-April unless carefully looking.

 

[mrblaine]

Very clean, simple, but I'm not sure my skillset is up to that level without thousands of dollars of tube bent under my belt.

We all have our priorities and projects all cost what they cost.
You can save money paying a pro by doing it yourself as long as you factor in time, purchased equipment, and waste.
You can pay a pro to do the hard stuff and save purchasing equipment, time, and wasted material.
You can save money by doing it all yourself and being far less picky about how it turns out.

I've run into the last one more times than I care for. I was asked to install a set of Genright high lines. When I gave them the minimum number it would cost, I got the "fuck that, I'll do it myself". When they brought the rig back later for some other stuff, it was the single most terrible install job I've ever seen. Had I been able to hold my nose long enough to do the job, I could have done it for 250 bucks.

They didn't attach the inners to the outer, they just sorta got them close and used long sheet metal screws to sorta hold them in place. There was literally a gap at the front you could stick your fingers through. The quality went down from there.

 

[macleanflood]

Osh Cut will bend tube to your specifications. If you can draw it up in CAD then you can essentially hit print...

-Mac

 

[mrblaine]

Osh Cut will bend tube to your specifications. If you can draw it up in CAD then you can essentially hit print...

-Mac

If you can draw it correctly in 3D CAD, you can just as easily bend it.

 

[macleanflood]

If you can draw it correctly in 3D CAD, you can just as easily bend it.

Agreed if you have the bender or want to invest in the bender and dies. Just saying if you want a custom cage and don't want the expense of the tools then it's an option to consider.

That and bending tube takes practice and skill...I could probably figure it out but I'd probably burn up a tube or two figuring out where to start a bend and how much spring back the machine had.

-Mac

 

[mrblaine]

That and bending tube takes practice and skill...

And 3D CAD doesn't?

 

[mrblaine]

figuring out where to start a bend and how much spring back the machine had.

-Mac

The bender has zero spring back. Tube on the other hand, does and how much depends on the CLBR, type of tube, and alloy.

 

[macleanflood]

And 3D CAD doesn't?

Fair point. Most folks already have a computer and the software is pretty accessible. More so (imho) than a $2000 tube bender and $1000 die set.

I definitely need to spend more time learning CAD... probably solid works.

I've got some things I'd like to model up and make...from power pods that tuck up under the angle on the rear of the tub left and right of the doors...

-Mac

 

[mrblaine]

Fair point. Most folks already have a computer and the software is pretty accessible. More so (imho) than a $2000 tube bender and $1000 die set.

I definitely need to spend more time learning CAD... probably solid works.

I've got some things I'd like to model up and make...from power pods that tuck up under the angle on the rear of the tub left and right of the doors...

-Mac

Where are you going to get the TJ interior file to model the cage against?

FYI- I don't have a tube bender. I can borrow one tomorrow. I can't do that with CAD, or the skills to run it.

 

[macleanflood]

Where are you going to get the TJ interior file to model the cage against?

There were some folks...might have been in the 3d printing thread... talking about scanning parts of the TJ.

3d scanner prices have dropped a lot recently and accuracy has greatly improved.

I'll bet if you scanned a number of TJs from 97 to 06 you'd find plenty of discrepancies...as you probably know well.

All I'm saying is everyone's situation is different and you have options.

-Mac

 

[sierradmax]

Luckily, my Jeep club has a JD2 bender, converted to hydraulic, w/ a 1.75" die. And club members are welcome to use it.

 

[mrblaine]

All I'm saying is everyone's situation is different and you have options.

-Mac

All I'm saying is you may be a tad overly optimistic about how easy it would be to draw up a cage in 3D and get the tubes bent to where they resemble anything able to be used.

Here is how it actually works and it is identical to the stupid 4 link calculators. All your efforts to get the parts back from the CNC tube bender are for samples to see how they fit and then get adjusted in CAD for a set that fits well enough to almost work. Like the calculator, the first set is a suggestion for a starting point to see what needs to be changed.

 

[Wildman]

Luckily, my Jeep club has a JD2 bender, converted to hydraulic, w/ a 1.75" die. And club members are welcome to use it.

Good luck with your project and look forward to seeing what you come up with. Since I don't own a bender & don't plan on buying one I'll have to farm out this part of the project.

 

[GoldenGorilla]

Where are you going to get the TJ interior file to model the cage against?

FYI- I don't have a tube bender. I can borrow one tomorrow. I can't do that with CAD, or the skills to run it.

3D capture and 3D modeling are moving fast and quickly becoming fairly accessible technology. Advances in 3d scanning are happening pretty quickly. For $500 you can get into it with something like this.
https://www.revopoint3d.com/products/infrared-laser-3d-scanner-inspire2

I don't disagree with your point about the first version being essentially a rough draft.

 

[sab]

3d scanner prices have dropped a lot recently and accuracy has greatly improved.

3D capture and 3D modeling are moving fast and quickly becoming fairly accessible technology. Advances in 3d scanning are happening pretty quickly. For $500 you can get into it with something like this.
https://www.revopoint3d.com/products/infrared-laser-3d-scanner-inspire2

Have y'all read my recent post in the 3D printing thread lamenting the difficulty of 3D scanning? It's here:

Post in thread 'What have you 3D printed for your TJ?'

Apr 21, 2026

I'm jealous, k1n3k! I've dabbled with scanning, and it's frustrating to get a scan and then turn that into a solid model. I've started and stopped twice now, and I've gone from a $500 scanner to a $2,000 scanner. I've also paid for a year of Fusion and bought a $1,500 version of Geomagic Wrap. So, I've got over $4k invested and nothing to show for it, yet. I currently lack the time required to develop a workflow.

• sab

• 
• 
• 

I'm pretty handy in the 3D modeling world, having starting in the mid-90s with ProEngineer. 3D Scanning, and the associated workflow required to turn point clouds into solid models, is something new, and it's currently kicking my butt. There's a steep learning curve to learn 3D modeling, and then the curve for turning 3D point clouds into something useful is equally steep.

I've also fabbed a tube chassis from scratch (SAE Mini Baja, 40 years ago.) I'd compare the learning curve for fabbing a role cage without CAD to that of learning either 3D modeling or 3D scanning. Learning both of those? I'd put that in the "harder then fabbing a tube chassis without CAD" category. It'd be easier to do it the "old-fashioned way."

And don't even get me started on how crappy the $500 scanners are. Well, at least the one I bought was, even though it had great reviews, which had to be fake. It was a 3DMakerPro Moose and was absolutely useless to me. I think I got one poor quality scan out of it, despite trying about a dozen times. The Creality Raptor Pro with their Scan Bridge is much better, but at nearly four times the price. The biggest difference is that the Moose was infrared only, while the Raptor Pro adds blue laser. You have to use adhesive targets with the blue laser, but I can actually get a scan to work with it, unlike the Moose.

Hey, @Fluxor, you're in this thread - did you scan the TJ model you use for your work, or did you get it through SEMA?

 

[sunnysideup_lj]

I have used scans to model for quite some time. We had a FARO arm at my work to scan UTV’s to make parts. This worked really well, but was very time consuming to get a scan converted to be useful for modeling.

Whenever I had a tube project that required multiple bends on different planes I’d resort to the angle finder and taking notes for production on the CNC bender.

I have the SEMA TJ frame file. It has been helpful when designing my bumpers and body mounts. Mostly accurate from what I’ve seen.

 

[mrblaine]

3D capture and 3D modeling are moving fast and quickly becoming fairly accessible technology. Advances in 3d scanning are happening pretty quickly. For $500 you can get into it with something like this.
https://www.revopoint3d.com/products/infrared-laser-3d-scanner-inspire2

I don't disagree with your point about the first version being essentially a rough draft.

My uncle did large construction projects. One truth was never violated. The longer you spend away from doing the actual work, the faster you get at it when doing estimates.

In the same vein, it is the rare highly skilled individual who can remember the time when they weren't and stop expecting others to do things as easily as they do. I can do shit in my sleep that others struggle with. I can see and solve problems faster than most I know. I'm smart enough to also know it is a fool's errand to expect the same from others and tolerance must be in play.

In other words, I don't care if they are giving away the best scanners in the world for free, it still requires experience and skills to convert that to something useful and to simply say draw it and hit print is no less retarded now than it was 20 posts ago.