Pirate 4x4 banner

Roll Cage 101

132K views 130 replies 79 participants last post by  DRH.1979 
#1 · (Edited)
Roll Cage Design 101

Roll Cage 101

I got inspired to write this article after seeing countless ill-conceived designs and poorly executed fabrications on the internet. This article is intended to be a guide to the basics of cage design. I don’t claim to know everything in the world about building cages and chassis, and I am not a metallurgist, but I am a CAD designer, fabricator and welder. I am always interested in learning more, so I welcome corrections and addendums to the following guidelines. I would like this thread to be a sticky, and I will edit this post, incorporating any good suggestions that are made.

Material:

There are three basic choices for material.

Mild steel tubing is made from sheet that is rolled and welded. The alloy is typically 1010 or higher. This material is not as strong as the others, but is totally acceptable with a proper design. It is even preferred by some for its tendency to bend before breaking.

DOM steel tubing is manufactured the same way as mild, including the welding. The alloy is typically 1018 up to 1026, the higher the number, the higher the carbon content. DOM means Drawn Over Mandrel, which “trues” the tube and hides the weld, giving it more accurate dimensions. DOM is about twice the cost of mild, and almost as much as 4130.

4130 chromoly steel tubing is a true seamless tube, with chromium and molybdenum added for strength and light weight. 4130 is very expensive and is used most often in big budget builds. It requires heat treating after welding to achieve maximum strength. Some say that if a 4130 chassis is not heat treated, it is no stronger than the other steels. 4130 suspension components should definitely be heat treated.

The size of tubing is determined by the weight of the vehicle, or the class it will be raced in, along with other factors. In general, it would be wise to use a minimum of 1.5 x .120 for lightweight vehicles like sand rails, 1.75 x .120 for mid sized vehicles like pre-runner trucks or Jeeps, and 2 x .120 for the heaviest vehicles like trophy trucks or huge 4WD buggies.

Bending the tubes:

The first rule of bending tubing is that no deformation is allowed. The bend must be smooth with no scoring or ovalizing of the tube. Do not use a pipe bender to bend tube, because the dies do not fit correctly. The HF “kinker” is infamous for a terrible bend on tubing.

Notching the tubes:

There are countless methods of notching a tube so it will fit tightly to another tube prior to welding. This is also called a fish mouth. The most common way to notch is with a hole saw, which is often done in a Tube Notcher tool. Cheap ones are available from HF, and high end models are available from several manufacturers. There are methods of notching on a mill or lathe, and there are also expensive dedicated machines for notching that use end mills or abrasive belts. A proper fitting and tight notch are extremely important for a strong weld joint.

Designing the cage:

There are many design rules that should be followed whenever possible.

The most important rule is to make triangles, not squares, and especially not artsy-curvy designs. Look at any structure in the world, whether it is a bridge, a crane, or a cantilevered sign, and you will see triangles everywhere. This principle should be applied as much as possible to a chassis or cage design. Every tube should be one leg of a triangle if possible. This is especially true of the primary structural tubes. The peripheral tubes like bumpers can be more “artistic”. It’s a good idea to build crumple zones into bumpers to absorb crash damage. It is better to crush a bumper than to damage the main chassis structure. A bolt on bumper makes repairs much easier than a welded on structure.

Bends are not your friends. Some bends are unavoidable in a design, but they should be minimized, not maximized. Even a perfect bend is weaker than a straight tube. Bends should never be mid-span, or unsupported. The apex of a bend should be a node point or junction for at least one other tube, and gusseted unless several tubes meet at the node. An example of a node is the center of an “X”.

It is advisable to gusset corners, especially when building a bare minimum cage. This can be done with triangular plates welded into the corners. A stronger method is to weld a 6-12” tube diagonally in the corner, similar to the letter A.

“T” junctions are called a dead tube junction, as one tube dead ends into another. This should be avoided whenever possible, because the dead end tube will bend the other one when the loads are along the dead tube.

“A” pillars should not be leaned back too far, unless a second A pillar is added to triangulate it. Otherwise it can collapse into the passenger compartment. The B pillar will be strongest when near vertical. It is always safer to double up on the A and B pillar on heavier vehicles. All cages benefit from a vertical tube in the windshield area. An inverted “V” like this /\ is even stronger.

The B hoop should have an “X” built into it, or at the very least a diagonal or a V. If the A and B hoops are inverted U shapes, the “spreader” tubes that go between them should intersect the apex of the bends for greatest strength, and they should be straight. The roof area should have a V or X built into it, depending on overall design. The B hoop needs to have rearward supports, typically at a downward 45 degree angle. If the B hoop does not have an X, then these tubes definitely should.

No tube should EVER terminate into sheet metal, like a floor or firewall. The A and B pillars should pass through the floor and weld solidly to the frame rails or tubes. If necessary, it is acceptable to weld a plate to each tube, above and below the sheet metal, and use four bolts to connect them together.

Most exo cages seem to be designed to protect sheet metal more than the vehicle occupants. Which do you value more? If an exo cage is a must for you, try to incorporate as many of these design features as you can, especially in the cab area.

Some say that any cage is better than no cage. This may be true in some cases, like flopping over at zero mph. But what if a failed hill climb attempt results in looping out and a triple endo down the hill? What if an off camber slip results in six barrel rolls down a rocky slope? What if one of these tumbles comes to a sudden halt against a large tree or rock? It is in these worst case scenarios when you need a properly designed cage and chassis, fully triangulated, and expertly welded. A marginal cage could collapse, doing more harm than good.

If you are not 100% sure that you can make a strong high quality weld, leave it to a professional. In general, TIG welding is considered superior to MIG welding, but a proper MIG weld is completely acceptable and just as strong. Tube splices and repairs should always be sleeved for strength and rosette welded, never butt welded.

The more you integrate these “rules” and suggestions into your cage, the stronger and safer it will be.
 
See less See more
#86 ·
If you plan to rub your rig along the roof line against rocks ... or at least want to prepare for that eventuality (driving out of a partial flop, for example, or sidehill on one side of a notch while you scrape through on the other side) then the smoother line of an A-to-C continuous bent part is probably better than an A hoop and a B hoop with stringers and gussets.

I'm building with the thought that every surface of my structure is basically a ski.
 
#88 · (Edited)
I'm aware of more stringent requirements for cage construction for SCORE than for Hammerking/KOH/Ultra-4 (SCORE, I believe, specifies TIG?) and I'm sure there are additional requirements in other motorsports. Unfortunately, I know zero details about this. The vast majority of the cages for trail rigs I know of are MIG welded, and the same is true for KOH racers. The majority of the comp chassis designs appear to go A-to-C.

Here are a lot of pictures of rock racers with cages that are almost all identifiably A-to-C single hoop constructions with an added B pillar and a possible wing support for the A pillar.

http://www.pirate4x4.com/gallery/main.php/v/vorra-3-10/
 
#89 ·
I believe there are too many variables left out to simply say "smooth A to C" is stronger or weaker than "hooped A, B, and C".

Personally if its an open top or buggy i prefer smooth A to C simply because it acts as a slider (as mentioned above). A full roof vehicle? I say build whats easiest to fit in there.
 
#93 ·
No lack of tube in THAT cage!

A couple of thoughts, though... It looks top-heavy with not only an X in the roof but a CROSS and an X put together! It's literally an 8 point star. I'm not so much worried about the weight when I say top heavy, I just think it's "busy". There are unecessarily tight angles in there to weld inside of. Acute angles suck for welding. I wish I'd avoided some, in my build.

Next, and probably more importantly, the lateral shear strength of a cage is greatly facilitated by having angled tubes behind the seats. This can be in the form of an X or a V. The central pillar in that build doesn't do anything for shear other than resist racking by the strength of the welds at its ends. Angled tubes, on the other hand, resist shear by taking a tension/compression load.

Although the above are basically nit picking points (I'm sure this cage is PLENTY strong to lay over on it's side or even go completely on it's lid), I have one concern that is a little more serious.

The B pillars are just ahead of the seat head restraints. So is that central pillar. This is a potential for major "headache". Do you see how the shoulder bar wraps around behind the seats with some bends? That bar is probably in a plane that would be better for the B pillar, relative to the seats. It looks like the position of the B pillar was chosen based on the body lines, so some degree of parts-compromise is necessary. Do you follow the seat lines for clearance on your head, or do you follow the door lines in the body? There are pros and cons to both.

Another area that you can get into some trouble with is head room. You really have to pay attention to this and think about it ahead of time. If you build a cage to fit inside the space defined by a removable hard top or if you just use the vehicle profile with hard top to draw your lines, then you're already invading into precious head space and every fraction of an inch counts. Consider making enough space that you can wear a helmet and still have enough space for a fist between the top of your helmet and the overhead bar.
 
#94 ·
Belly Doc that is the most constructive and helpful post I have ever received from a forum. Thanks! I had not noticed the "B" pillar but I do agree that it needs to be moved back a bit. I will also not be installing the center post but I will be using an "X" design behind the driver & passenger to help with the shear force. My TOP concern is my children so I will also be using a "X" or inverted "V" on the "C" pillar behind them.
 
#100 ·
SFI sells NHRA cage specs for $35 plus the ride
every tube has a number. makes it easy to discuss placement
25.4 is a 3600#, mild steel, mig welded cage that NHRA says is required 8.50 to 7.49sec.
you can learn a bunch by looking at what they do
this sport is big enough to emulate what SFI has done for fast NHRA cars
 
#103 · (Edited)
Let's bring this back.

Since everyone seems to have a different opinion on what they think is right, why not start posting pics on what you think is right? I think some visual references would be great.

Now for my question, (I'm sure it's somewhere in this thread, I probably just missed it) do you think it is okay to re-use the B and C pillars on a Jeep YJ and just add on a new A pillar with lots of triangulation? I'm wanting to build my roll cage but I want to do it the right way. After all, it's me inside!!!

EDIT: Here are some pics I found on the internet, give me opinions. Obviously I would not be re-using the stock bolt-in bar between the A and B pillars, that would be a new weld in tube. Good, bad and ugly.

Thanks,

-Joe
 

Attachments

#114 ·
Why would you be using two different size tubes there? Nevermind, doesn't matter. I would say that it could only be stronger than cutting/notching/welding. Where a notched X intersection can get crushed, a pierced intersection seems like it might be better in that respect.
 
#116 · (Edited)
Poly mount?

So, what's the answer for poly mounting, then? Just stuff rubber biscuit style body mounts between your cage and the frame, using bolts to hold it all together?

What if you used those "can" style poly mounts from way upthread, in pairs, but then ran a solid tube between the pieces that were rising from them?

I'll edit in a minute and add a picture.

------

Ok, imagine this is a very small main hoop. Crossbraced corner to corner, with the crossbeam at the bottom. The vertical members curve inward and have poly mount eyes that use whatever sort of shackles you like to bolt to the frame rails. (Or get welded on, whatever.)



Thoughts?

------

Ok, after thinking on it again...

I suppose in a complete rollover, you're loading vertically from the top, and with the pivot points of the eyelets, you're still counting on the strength of the tubing in those curves at the bottom to keep the entire structure from moving downwards. (The bottom tubes pivoting around the eyelet bolt as the curves crush and bend in the center.)

So, ok, that probably doesn't work, then. What is the right answer?
 
#117 ·
It has been talked about before but headroom.

If you are building a cage and you are 5ft tall but your buddy who rides with you alot is 6 ft tall well its going to be a dangerous place for him.

This has got to be one of the most common mistakes I have seen on cages, bars to close to your noggin. It hurts when said noggin contacts tube, be it DOM HREW SCHD 40 still hurts about the same.
 
#118 ·
Well, yes. The picture is just to demonstrate the mounting style down at the bottom. To scale, that "main hoop" is all of 20 inches tall. I threw in crossbracing to show how it would be crossbraced. I'm not actually planning on making a roll bar that's shorter than my knee.
 
#120 ·
ilcylic-
Sure, lots of people mount their rollcages to the frame like that. Great way to insulate the cab from some of the noise and vibration that gets transferred through the chassis with a solid-mounted cab. You'd want to brace that hoop, and those mounts, differently, but I don't think that was the point? What's you're question though, specifically?
 
#121 ·
Well, so, in post 29 of this thread, a setup similar to that which I diagrammed above was displayed. And in post 30, someone had a, erm, vigorous objection to it.

I was wondering if the drawbacks to that design would be mitigated by the addition of the lowest horizontal crossbar in my diagram. But then I figured out that the answer is "no, not really".

So, my specific question is, what's the optimum means for attaching a cage to the frame of the vehicle? To be extremely specific, I've got an '84 Suburban with the entirety of the body about 6 inches behind the front seat cut off. I'm planning on putting a cage on/in this, along with a bed and some fenders & mudflaps to keep it street legal, and using it for "pickup duty" around town, in addition to being my 4 wheel drive off road toy.

I've seen a lot of back and forth on the subject of direct welding the cage to the frame on a frequently street driven vehicle, with the apparent consensus being that the rigidity of the cage, when tied to the apparently notoriously floppy Chevy frame, will lead to a lot of stress cracks that will of course significantly limit the safety of the cage in a rollover.

And that, while loquacious, is my specific question. I think. :)
 
#123 ·
A rollcage can absolutely be mounted to a frame with bushings in a safe manner. It's just all a matter of design/execution. Sorry, I'll post a more thoughtful reply sometime later this weekend. In the meantime, try searching race-dezert.com to find some good examples of how this is sometimes done in high-end prerunners.
 
#124 · (Edited)
A rollcage can absolutely be mounted to a frame with bushings in a safe manner. It's just all a matter of design/execution.
Isn't everything? :D

Ok. The explanation given for why this was a bad idea made sense to me, but perhaps that was just due to the specifics of that particular mounting setup.

I would be interested in hearing how to do it safely.

Other than that, I am generally inclined to go with what Niddle suggested above, and bend the hoop in far enough to hit the frame rails.

An actual to scale drawing of a couple potential B pillar hoops here:



Though now I'm thinking I might want a bar bridging the stuff at the bottom, too.
 
#126 ·
Something like what you posted above (124) would be fine, and even better if there was a horizontal tube spanning the bottom. That could be another tube, tied into the rollcage (at/near the bushing mounts?), or a factory crossmember. You just want to avoid mounting tubes (especially on bushings) in ways that will post twisting loads on the framerails. For instance, you wouldn't want to hang a bushing mount too far off the side of the framerail, mounted so the bolt would be parallel to the framerail. On the other hand, if you mounted all the bushings along the top of the framerail with the bolts oriented so they would be parallel to the axles, that wouldn't really leave the bushings free to rotate in any direction that would cause the rollcage to fail. Along those lines, another thing to consider is whether you are using bushings in order to be able to remove the body from the chassis, or just to reduce noise and vibration. If you're not concerned about removing the body, then it's even easier to orient the bushings in ways that would reduce flex. There are so many different ways to do something like this, it's not even possible to list them all, but it is certainly something that you CAN do safely. Don't listen to anybody that tells you otherwise.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top