Another RX7 in the mix

RX7Club has too much traffic really. Stuff just gets lost in the shuffle, and 95% of the people are just bolt-shit-on, let-other-people-work-on-my-car or tire kicker types anyway.

I note for the record that all my experience with 1st gen cars is roadracing (though there's close to ten years of that) so while I've tried to pay a lot of attention to where rally stuff is different and think I've figured that part out, well, grain of salt and all that.

Adam: I don't know if you remember Leonhardts' organizational capabilities, but I asked him about your paper and his eyes grew wide and he responded with "Uuuuhhhh.... Somewhere?" He said he would take a look around for them, but wasn't going to promise anything.

I noticed in the google photos you sent me that what you did with the trailing links is what Pete was talking about with the 4-link suspension and a Panhard bar. Am I right? I'm guessing the Corolla had similar issues to the 7 with the shitty, binding rear suspension.

Also, how did you go about constructing your cage in the car? My plan has been to cut holes in the floor to drop the ends of the cage through, weld it together, and then lift it up and box the ends. Does that sound about right?

Daniel: That's very true. Often times I will go on the technical threads to laugh at the problems and questions some people have posted. There are still a few here and there that are great knowledge pools (Jeff20B for example).
And even though road racing isn't rallying, it is still racing and there is a lot of knowledge that can be shared between the 2. My racing experience is mostly with Autocross and Formula SAE (student team, purpose-built autocross cars) but I don't feel like a complete 'tard most of the time when discussing other forms of racing.

Everyone: Finally, I'm still unsure about how to go about the rear strut towers. I'm not going to be able to afford nice long coilovers in the time I'm building my cage (or for a bit of time afterwards) and am staying with the stock spring and shock for now.
Tangent: In all honesty, I'll probably be autocrossing this car more than rallying it until I'm done with school due to the financial and proximity constraints involved with rallying. So greater suspension travel and ground clearance aren't vital ASAP.
So, back to rear strut towers. I was planning on boxing the rear struts much higher to accommodate for having longer suspension travel in the future and leaving the stock towers inside said box. When the time comes for coilovers I'd cut a section out, remove the stock suspension, and then already have the cage and boxed strut tower in the correct place to add the new coilover.
I guess I just wanted to run this by someone who would have a better idea about the do-ability of this plan than I do.

---

"The more I learn, the more I realize I don't know and have so much more to learn." - Claude Rouelle, Optimum G lecture June, 2011



Edited 1 time(s). Last edit at 01/13/2012 01:40PM by MeCalledEvan.

i use rotarycarclub.com a lot. Its for all the people who abandoned Rx7club after they were bought out and got flakey.

---

First Rally: 2010 First RallyX: 2004 (a bunch)
Driver (0), Co-Driver (7)
Organizer (3), Volunteer (3)
Cars Built (2.5), Engines Blown (2) Cages Built (0) # of rotations (3.5)
Last Updated, Apr 9, 2023

Quote
Gravity Fed
i use rotarycarclub.com a lot. Its for all the people who abandoned Rx7club after they were bought out and got flakey.

Sounds better. I'll have to check it out. Thanks!

---

"The more I learn, the more I realize I don't know and have so much more to learn." - Claude Rouelle, Optimum G lecture June, 2011

My view on the rear towers is do it once and do it right. If you can't afford to do it now, then wait until you can afford it. I don't see how you'll be able to build the new towers and leave the old towers there and useable. In reality though the long rear shocks aren't that expensive...maybe $100 each more than the KYB's... it's worth getting these, even if you have to save for a couple of extra weeks/months. I presume John is recommending the Bilstien B46-1xx coil-over shocks.

Rally car builds are never fast or cheap. I guy I know who's buildinga Civic just todl me his budget is looking like 4x what he planned... he must be a government project planner :-)

-Martin.

Quote
MRWmotorsports
My view on the rear towers is do it once and do it right. If you can't afford to do it now, then wait until you can afford it. I don't see how you'll be able to build the new towers and leave the old towers there and useable. In reality though the long rear shocks aren't that expensive...maybe $100 each more than the KYB's... it's worth getting these, even if you have to save for a couple of extra weeks/months. I presume John is recommending the Bilstien B46-1xx coil-over shocks.

Rally car builds are never fast or cheap. I guy I know who's buildinga Civic just todl me his budget is looking like 4x what he planned... he must be a government project planner :-)

-Martin.

Martin, that's usually how I go about building things as well. But new shocks, period, are out of my price range at the moment. I'm on a college student budget, and the only reason I'm caging the car now is because I'm taking an independent study this quarter to build it. I can use part of my "college fund" money I've saved up to buy materials needed for the cage, but not new $170 shocks + price of coil springs as well.

No matter how badly I want them buying 2 of those shocks for the rear suspension is about the grand total of what I make in a month. Perhaps I'll focus on the front end of the cage first and I might have enough for the shocks by the time I get to the rear struts.
Ideally, the cage is supposed to be done by the end of the quarter (in March), but I honestly don't know if that will be possible. To receive credit for it though, I will need to be well underway with it at the least.

I know that building a rally car isn't cheap, and probably isn't practical for me, but I'm trying to build one anyways over time to learn the process of it. I'm a vehicle engineering student and I want to get into race engineering after I graduate, so building this is more of a learning experience than anything else for me.

---

"The more I learn, the more I realize I don't know and have so much more to learn." - Claude Rouelle, Optimum G lecture June, 2011

You really want to build the 4 link boxes and turrets when you are caging the thing so that you can tie it all together.

---

"Talk about drugs. Driving a car like that, going that fast, it’s like all the drugs at once." - Tommy Byrne

"Now, Pinky, if by any chance you are captured during this mission, remember you are Gunther Heindriksen from Appenzell. You moved to Grindelwald to drive the cog train to Murren. Can you repeat that?" - The Brain

That's what I said I wanted to do: build them with the stock strut tower/mount still in it so that the cage is ready to have longer coilovers installed when I can afford to buy them.

---

"The more I learn, the more I realize I don't know and have so much more to learn." - Claude Rouelle, Optimum G lecture June, 2011

Building the towers is a hell of a lot cheaper than building a cage, even if you're doing it yourself.

Something else to consider: Imagine trying to fabricate that stuff AFTER the cage is in the car. Working in caged 2-door cars sucks.

---

Pete Remner
Cleveland, Ohio

1984 RX-7 (rallycross thing)
1978
Silence is golden, but duct tape is silver.

I've been way busy so this has gone on a bit...har harumph..
Kid, the way you build your boxes and towers once---before you think of cage is leave some extra holes in the forward ends of the box or mounts.

The cooolest car in this parsec of the Galaxy, the unconquerable Ford Escort, did that so they could: pill the front upper and lower link bolts, slide 'em down and shove the bolts in, the bang on some shorter shocks and struts and go..instant tarmac car...

If your saying you'll be mostly Otter-crossing, then don't bother with the cage...and don't bother with shocks or springs.. Take a tube, hammer the ends flat and drill a holt in one end and for the pin mounts just zot a M10 bolt on the other end, shobe it thru the hole and presto you're done..
Trust me, every other forum I've looked at Otter-crossers are using springs easily 200 to 300% stiffer than tarmac rally guys on flexie flyer stock unstichwelded, uncaged cars so OBVIOUSLY the hot ticket is no suspension movement, just body shell torsional twist.

The flattened bar method is a lot stiffer and a lot cheaper than buying suspension that works....

Now there's one loon here driving a extremely well prepped car that before he built it he was telling me he wanted to build a doriftosis car, and evidently I told 'em this "DoooooooD, built it like a rally car with good cage, and stitchie welds and good suspension. Then when there's a driftosis day, gloovy; when there's a track day, go blaze; if there's a T-bird, cool; and if theres a play in the woods day, well you're set. A well set of rally car can certainly be drifted, or track day-ed and it doesn't make any difference, but a dorifto car is only a dorifto car and an optimise road race car is only good for that..."

So 2 years later he reminded me of that... Sometimes forgetting sumpin brilliant means you have the fun of being brilliant again...

So let's review:
You gonna be a broke ass college student and go to Alchie U aka WWU./
You're probably going otter-cross
so a) forget the cage NOW
b) forget suspension, a solid tube will do
c) by time your done the cage will be dead anyway.
d) but not a loss; a gain. Cause then you can buy a $200-400 240 and have MORE FUN!

---

John Vanlandingham
Sleezattle, WA, USA

Vive le Prole-le-ralliat

www.rallyrace.net/jvab
CALL +1 206 431-9696
Remember! Pacific Standard Time
is 3 hours behind Eastern Standard Time.

Quote
john vanlandingham

So let's review:
You gonna be a broke ass college student and go to Alchie U aka WWU./
You're probably going otter-cross
so a) forget the cage NOW
b) forget suspension, a solid tube will do
c) by time your done the cage will be dead anyway.
d) but not a loss; a gain. Cause then you can buy a $200-400 240 and have MORE FUN!

Words of wisdom John. Thank you for going through and reading everything I've been saying and giving me some valuable feedback.
And as for the hole in the box/the Escort: that's a good point I hadn't thought of.

So, I would scrap this whole idea to cage the car this winter but, I'm in a bind ... I've committed to doing this independent study for school already. I should have started this thread a couple months ago when I was writing the proposal to do it. It is now a part of my class schedule, and I am supposed to design and build a rally car roll cage for my Mazda and receive 6 credits for it. I can't drop the course because that makes me no longer a full-time student and I lose financial aid.
So... it looks like I am still going to be building a cage! Wise decision or no.
The upside is that I will be learning. I may not end up with a bad-ass balls-fast rally-mobile at the end of the day, but I'll probably have an idea of what would need to be done to eventually morph this car into something race-able throughout the next few years.
Plus I have you guys to go to for advice here and there when I'm in over my head again... I hope!
But, point being, I am literally incorporating this car into my education. And I'm ok with that. And who knows... maybe I'll pull it off and it will turn out alright. I'm not quite as dumb as I look!

What you told that kid about building a rally car to be capable of any type of racing gave me an idea about what to do for the rear strut towers. --- I can box them. I can also use the stock spring and shock in them. I can fix the rear suspension around them (though for a lowered car, not a raised one), since this will lower the rear of the car and give me a ton of negative rake. But since I'll mostly start with otter-crossing, the fix can be to cut the front springs and lower the entire car. As unholy as cutting the springs seems, this should give me a cheap shitty car that I can try to dodge cones with. A quick fix until I can afford the nice longer coilovers this summer, get those, and remake some trailing arms to work with the different geometry of a higher ride height.
That way, I can have the car be able to autocross, go to a track day, or play in the mud. Kind of like the unconquerable Escort, just a lot less good.

---

"The more I learn, the more I realize I don't know and have so much more to learn." - Claude Rouelle, Optimum G lecture June, 2011



Edited 1 time(s). Last edit at 01/13/2012 11:07PM by MeCalledEvan.

Evan

Build the cage. Use the correct RA tubing or better. Build the cage so it meets the SCCA IT specs but do it so that it can be upgraded to rally spec at a later date. ( I seem to recall that SCCA was a wee bit restrictive on the number of tubes and placement for IT road race.) It will be less money out now but the bones will be good. You will then have several venues the car can be compliant to run or sold to later.

Make sure the cage will meet RA requirements when designed and built. Feel free to ask for help as you go. I'm the western RA tech supervisor.

Gene,
That's a great idea. The 2 differences between RA and SCCA cages that I can find are: 1) I won't be able to weld a bar along the bottom of the door sill and 2) I'm limited to between 6 and 8 mounting points.

Thank you. I'm going to start measuring the interior or my Mazda today and entering nodes into a CAD model. I'll keep everyone updated.

Evan

---

"The more I learn, the more I realize I don't know and have so much more to learn." - Claude Rouelle, Optimum G lecture June, 2011

As I understand it, your biggest conundrum is getting the rear cage geometry right... And if you don't do towers now... You'll end up with shorty rear suspension...

If you're out of cash to buy the long rear shocks, can't you just weld up an extension for your stock junk? These don't take side load like a strut right... You said you're not going to actually rally it for a while... Build your towers to the lenght you want and weld up a tube with a top pin on one end and a bolt on the other... Add that to stock shock and go crush cones when your done.

Why not?

Matt

text of my roll cage paper [written in 2005]. sadly the photos dont carry in, but here's a link to where most of the photos should be.roll cage photos

Roll Cage Fabrication

Adam Crane

WWU Vehicle Research Institute

Copyright © 2005 SAE International


ABSTRACT

The purpose of this paper is to describe the necessary steps to produce a roll cage legal for competition in rally events in the United States.

INTRODUCTION

A roll cage is a structure that protects the occupants of a race car in the event of a crash. This is accomplished by a structure of tubes that surround the occupants and reinforce the car’s body-work. The historical purpose of the roll cage was to protect the driver only in the event of a roll-over accident. This was often done with only a roll hoop; a tube bent into an upside-down U shape that was positioned behind the driver’s seat, and extended above his helmet. This is now called a roll cage’s main hoop. Rear stays were added to keep the hoop from tipping if a car overturned and skidded on the hoop. A-pillar bars accomplish the same task, while protecting the front of the driver compartment. Door bars protect the driver from intrusion when the car doesn’t flip, but slides into an object or another car. The cage evolved to reinforce suspension mounting points and stiffen a car’s chassis, while providing more protection for the driver.



Figure 1: A roll cage diagram from Rally America Rules. Note the Main Hoop, sill bars and gusset areas.

Constructing a roll cage is complex process. There are two main operations involved, design and fabrication. The design of a roll cage takes into account safety, rules, and suitability. Fabrication is the act of producing the design. Cage fabrication involves 3 main skills: Bending, notching, and welding. The bending notching and welding are done in a sequence that allows the best fitting cage to be made.

design

The design of a roll cage depends on a number of factors. Competition intent, sanctioning body, rules, cost, safety and ergonomics are all considered. The intent to race a certain class of car, a performance sedan for example, may enable you to pick and choose the sanctioning body with the closest competition or cheapest entry fees. Sanctioning bodies often have agreements allowing cars from other series to compete with minor modifications. The rule books for different sanctioning bodies will allow a cage to be built legally for a large number of series. Design the cage based on the rule book. Often there will be relatively few choices to make if the cage is to be legal.

Initial Choices

For the purpose of this paper a roll cage was designed to meet the Rally America 2005 rules, in the Group 2 class.

The rules state what elements must be used, the type of tubing, construction method, car attachment method, occupant safety standards and reinforcement areas.

Table 1. Roll Cage Information

COMPETITION INTENT

Performance Rally. Production car based, timed racing,

on unimproved roads

SANCTIONING BODY

Rally America

RULES/CLASS

Rally America 2005/Gr 2

CAR

1986 Toyota Corolla GT-S RWD, 2-door coupe

TUBING

1.5in dia X 0.120 wall DOM steel a-513

Suitable for cars weighing 2500-3500lbs

WELDS

MIG, Miller 175

with 0.030 wire



Design Choices

A good cage design should not repeat the poor designs choices of other competitors. Ask what features of a cage other competitors find useful or awkward. Try to locate data on crashes and cage performance. Also run new ideas past a local cage builder or better, a licensed scrutineer, who will be the one judging the safety of the cage before an event.

The roof of the cage is left free for individuals to design. A diagonal bar was put across the roof with a slight bend to meet the windshield bar, and maximize interior space. It is important to consider the seating positions of the driver and co driver here. The rules state that there should be a minimum of 2in between a helmet and the cage. Often the co driver will sit lower and to the rear of the driver, to help with weight distribution. Make sure the diagonal is a maximum distance from helmet positions.



Figure 2: Diagonal roof bar. Note slight bend near main hoop junction.

Gussets are required on both sides of the windshield bar in the 2005 rules. Four gussets must be made, each strengthening one side of the windshield bar, A-pillar bar junction. The gussets should extend between 2 and 3 inches onto the tubing in each direction. There are 3 different acceptable gussets; the taco, the plate, and the tube. A taco gusset is a sheet of metal the same thickness as the tubing that is bent 180 degrees to resemble a taco shell. This shell is then trimmed so that the edges enclose a tubing junction, adding strength to the joint. The taco was chosen as it would be the easiest to incorporate into a new cage and presented the least injury potential in a crash. The plate gusset is a 1/8th inch plate of steel that is fit into the smallest area of a tubing junction, like a pizza slice joining the two tubes. A plate gusset presents a sharp edge to flailing limbs. A tube gusset is a ¾ inch tube that is notched to extend from one tube to another within 3 inches of the joint. This gusset would require ordering an additional tubing size and be the most difficult to weld properly.



Figure 3: One of 4 taco gussets on the driver’s side of the windshield bar. In its final position, note lack of welding clearance.

Rally America leaves the competitor the choice of an X-brace or V-brace in the main hoop. In the Corolla, because of the long doors in the coupe layout, an X-brace would be redundant, with the rear diagonals so close to the main hoop. A V-brace was chosen to ease access to the rear of the vehicle, where the battery and tools are stored. The V- brace also saved approximately 4 feet of additional tubing.





Figure 4: V-brace. Note accessibility of battery box.

The rules require a Sill Bar in addition to any other door bars. This bar is intended to keep smaller stumps from breaking through a cars lower sill and impacting the driver or co-driver. According to the rules, the Sill Bar should be as close to horizontal as possible and within 6 inches of the sill. For the Corolla application, with its long doors, more bracing then only a sill bar was desired. It was decided that a light, effective X-brace could be made in the door opening to protect the occupants. if the Sill Bar was started 6 inches above the sill at the front of the door and sloped downward to the base of the Main Hoop, it would fulfill the rule requirements. The rest of the brace is an interrupted tube running from door-latch height to the base of the A pillar bars. This design meets the rule requirements and safety concerns, without a redundant bar along the strongest part of the car, the sill.



Figure 5: Sill bar with diagonal brace.

Additional bars were placed from the bottom of the main hoop up to the rear shock towers. These bars are intended to triangulate the rear suspension forces into the rest of the cage.



Figure 6: Triangulation bar from the shock tower to main hoop.

Design and Fabrication

A cage tube must be welded completely around every junction and gusset to avoid weak spots. A good weld can be accomplished with as little as 2.5 inches clearance. A great weld, which is the goal, requires approximately 4 inches.

The strength and safety of a cage depend on how closely it fits to the shell. So, to weld the cage where it is close to sills, doors and the roof, the main spider must be somewhat mobile. Main spider is a term used to describe the portion of the cage that encloses the driver. This includes; the main hoop, A-pillar bars, windshield bar, roof diagonal and door bars. The main spider of the cage is the most important part of a tight fit. Once it is tacked together it must be shifted to gain clearance.

Unlimited welding clearance can be gained at the expense of some body work, by removing the roof after the main spider is tack welded. The top of each junction can be easily accessed. Restoring sheet metal, roof, door and window gaps to factory appearance makes this method very time consuming to someone without bodywork experience.



Figure 7: Tight fit of door pillar bars and windshield bar.

A roll cage inside a car is useless unless it is firmly attached to the car. If a cage tube was welded directly to the shell, the heavy duty tube would tear through the thin sheet metal in the event of a crash. Because of this, the cage must have “feet” to reinforce the attachment points.

There are two ways to shift the main spider of the cage. One is to build clearance into the cage feet, where the cage attaches to the floor of the shell. The other shifting method is squeezing, stretching, or shoving the cage inside the shell. Shifting and squeezing a cage inside a shell as it is assembled can be accomplished, but differences in cars can make this more or less successful than intended. Instead, it is better design movement into the main spider’s feet.



Figure 8: The shifted cage, allowing gussets to be welded. Note clearance between the A-pillar and cage. The cage is shifted to the rear of the car, with the A-pillar bar dropped to the floor.

The first design choice that effects the fabrication is cage foot style. By selecting how the cage attaches to the car’s shell, you are essentially choosing how clearance is for welding is made. Depending on the planned cage attachment method, different references are also made for cutting and bending the main spider.

The first foot style is a plate that rests on the floor or sill of the car. This is most often used for the rear and diagonal mounts. The rules dictate the surface area of the plate in most racing series. In Rally America rules the plate is minimum 1/8th in thick and has a surface area of 3 times the tubing diameter (15 in). Ideally the plate will extend up vertical surfaces as well. The cage tubing should be cut right to the floor or plate. By drilling a tubing-sized hole into the floor where the cage will rest then lowering the cage from the roof of the car, generous welding clearance can be made. When welding is completed, the cage is lifted into place and the hole is covered with the large surface area plate, onto which the cage is then welded. Downsides to the plate method include the hole that must be made in the structure of the car, and limited placement options, as it is unadvisable to drill into the sill of the car where multiple sheet metal layers come together.

A box is the other common foot style. By cutting the base of the main spider 2-4 inches short of the floor or sill, the spider can be dropped and roof bars and gussets can easily be welded. A box made from 1/8 in thick steel sheet is then made to bridge the gap between sill and cage. The same surface area and vertical attachment guidelines apply to the box. It is not necessary to weld a bottom onto the boxes, only to weld the box walls to the floor or sills. When fabricating boxes, the plates are heated then bent once or twice to present rounded corners to body parts. The advantage of the box lies in its ability to even up odd clearances, and the fact that a box can be built to any location. Disadvantages include the fact that a box must be bulky, and it presents an injury hazard in a crash.

By combining the box, plate, and a stretch/squeeze, a tight fitting cage can be made. The tightest the cage can fit, is with the tubing set against the door pillar, half on the sill. In this position a plate is inadequate and full box unnecessary. The top edge of the box is a plate, set half on the sill. Walls are made to box the remainder of the plate to the floor and vertical edge of the sill. To gain clearance, the spider is squeezed with a tourniquet then dropped off the sill, to the floor of the car. Alternately, the cage might be turned slightly then one leg dropped while the diagonal leg is lifted. Shifting the cage to the rear of the car gains roof clearance and also allows the front box to be welded on the door sill side.



Figure 9: The hybrid box style foot, made with one bend and two side-walls, for the drivers side front leg.

Fabrication Considerations and skills

The Fabrication of a roll cage is a careful process. A person’s life might depend on the cage’s construction. A haphazard approach will lead to many difficulties and a poor safety device. Always keep in mind the consequences of your actions. Build a roll cage that you would let your mother race in.

As the cage comes together, imagine the first crash, and its aftermath. The best way to tell if a cage is bent or tweaked is to start with a symmetrical one. On softer impacts, distortions can make the cage appear bent, or straight.

Tube Bending

Tube bending must be done carefully. Tubing is heavy, expensive and hard to bend. To avoid kinking a mandrel bender must be used. Tubing is fed between two dies and a wiper. The dies, specific to a tubes’ diameter, maintain the shape and give the bend radius. Tubing should be clean and free of rust, to keep from damaging the dies or scratching the tube. If the tube is damaged or kinked, it should not be re-bent or reused.

Successful bending requires practice.

Most cage bends will be shallow angles with the exception of the main hoop which has two bends close to 90 degrees.

To make adjustments to a bend, mark the tubing at the leading corner of the die with a permanent marker and a punch after the first bend, before removing the tube. This will allow the tube to be repositioned in the bender exactly. There are very few benders that allow a tube to be bent back straight because the metal deforms as it is bent and further bending induces fatigue that can lead to weakness and failure.



Figure 10: L-shaped marks on tubing, referencing the bender’s leading die, with center punched mark for reference.

To make multiple bends on the same plane a single line should be made down the length of the tube. This mark can be used as a reference for further bends. Find a door frame or angled piece of metal to hold the tube against while using the edge to make the line.

To make multiple bends, for the A-pillar bar or main hoop, start at the floor and move upward; Make a mark on the shell where you want straight tube to be after the first bend. Make another mark after the second intended bend. A tape measure is used to transfer the distance from floor to mark 1 and mark 1 to mark 2 onto the tube. These marks become the starting points for each bend. The tube is inserted into the bender so that the first mark is at the leading edge of the dies. The angle to be bent can be either measured with a protractor or a hinged piece of cardboard. After the first bend is close to the final angle, the match marks are made and if possible the tube should be checked for fit.

The second bend can be made like the first, though you must be aware of the bend plane in this instance. The A-pillar tube will be rotated and bent on a second plane. To find this plane, rest the tubing in a position relative to its final position. After visualizing the bend, use a permanent marker to mark a straight line on what will be the inside radius of the bend. This reference line will be placed against the bender’s die, closest to the center point of the die’s radius.

Tube Notching

Tube notching is the cutting of a tube so that it mates with another tube with no gaps. A good deal of the quality of a weld is how well the pieces fit initially. Notching a tube to attach perpendicularly to a similar size tube is a simple process.

A hole saw of a similar diameter can be used to make the notch if the tubing is held stationary. Tubing-notchers are sold that will maintain a hole saw and tube’s position while also cutting at an angle.

When tubing meets at an acute angle, a different notching method is required. Most tube-notchers do not have clearance to notch more than 45 degrees. Templates, and hand cutting are required for smaller angles. Templates can be purchased or produced on a computer. I used the Tube miter program, from the Internet. With these templates, certain modifications were made to make grinding and welding practical.

Using templates to notch tube takes more time than using a notcher. When the template is wrapped around the tubing, marks are transferred onto the tube. The contour of the cut is complicated, but can be thought of as a 90 degree angle centered about the perpendicular tube axis, rotated to the desired notch angle. Cutting can be done with a 4 ½ in angle grinder or oxy-acetylene torch. Grinding is best done with a small diameter grinder, which can reach into the pipe better.



Figure 11: Printed templates for 40, 30, and 20 degrees, CCW from top right.

The X-brace between the rear bars was made from 2 pieces of tubing, instead of 3. Each diagonal was cut, notched, and fit with the other removed. The bar that attached to the passenger side of the main hoop was tacked into place, and then careful measurements were made using masking tape as a guide for the second tube. The tube that was to be attached to the driver’s side of the main hoop was marked, and then notched through the center, cutting the bar into 2 perfectly fit, notched tubes. This method can save time if done carefully, if not, one half of the notched tube will be too short, and must be scrapped.



Figure 12: Finished rear X-brace.

Certain cage junctions will not be simple. Rear legs and diagonals meet at an angle in the upper corner of the main hoop. In these places it is best to use a scrap as a jig to obtain the best notching angles. In complex corners, start with two notches at a 30 degree angle on opposite sides of the junction. Use the grinder to adjust clearances until a tight fit is achieved.



Figure 13: Junction above the co-drivers head, a difficult notching job.



Welding

A well designed cage that is poorly welded is not as good as a solidly fabricated, basic cage. Welding a roll cage should not be a novice learning project. There are many aspects of welding in a cage that can catch a novice welder. Upside-down welds, tight quarters, blind welds, welds to dissimilar thickness metals, welding on ones side or back, and welding with a mirror are all techniques that can be used. Begin on less complex projects, and then practice the welds on actual tubing at awkward angles to be sure of good welds.

Keep the following items in mind when welding the cage.

· Clean all welds; the coating on most tubing to prevent rust can introduce contaminates to the weld. Use an emery cloth to sand the tube, then wipe clean. The same thing goes for painted shell areas.

· Cover all wiring with a welding blanket, a single spark can short out a large section of a wiring harness, and be quite difficult to trace and repair. Sparks will also pit glass.

· Welds onto vertical surfaces must not exhibit any significant under cutting.

· As the air inside the tubing heats up, it can force its way through an incomplete weld and lead to aeration of the weld. Drill a small hole within the circle of tube that is to be welded or, circulate around to different parts of the cage and allow the tubing to cool slightly.

· Tack the tube to be welded in more than one place on each end before welding to maintain the tubing alignment.

· Be mindful to weld junctions before they are covered with a gusset or another tube.

· Try to avoid large amounts of welding at one junction, as the heat can “tweak” the cage’s alignment. As much as possible, weld the cage tubing in its’ final position to maintain alignment.

· When welding the cage feet to the shell, it helps to go in short bursts, allowing the thin sheet metal to cool for 4 or 5 seconds after 2 seconds of welding.

· The rules say that there can be no evidence of grinding on the cage welds; this doesn’t mean no grinding. A poor weld, or one with aeration must be ground out and re-welded.

· Tubing that is improperly notched or fitted will lead to gaps that lead to poor welds.

· Since welding melts metal, tubing often shrinks as it cools. This can lead to the legs bending inward. A trimmed 2 by 4 can be used as a wedge to spread the legs back to their final positions.



Fabrication Step by Step

Start by removing from the car’s interior, anything that will get in the way of the ideal cage. This includes: door cards, dash board, wiring, seats, carpets, and plastic trim from the door frame, A-B-C pillars and roof. The best cage will sit as close to the car’s body as possible, thus leaving as much room for the occupants as possible.

Doors and windshield can also be removed if no interference is foreseen. Removing the windshield is a major asset, as much of the tight fitting and welding will be at the front of the car. Without doors, access to the inside of the car is un-restricted, but doors can also interfere in unexpected ways. You will end up with a stripped shell.



Figure 14: Stripped shell with seat.

If a similar driving position to stock is desired, it can be helpful to make head and body (elbow, knee, shoulder, and hip) reference marks on the shell. These reference marks will make sure that cage elements avoid the driver’s helmet space.

To begin with, the best way to visualize the cage is to make one out of masking tape. Pick the location of the main hoop base and A-pillar and rear diagonal bars. Stretch tape to get the main spider (A-pillar bars, main hoop, and rear bars). From here door bars and diagonals are easy to place. Use toilet paper or paper towel rolls to show the diameter of the tubing. This will give you an idea of what bars go where or where bars can’t possibly go. You will also get a big ball of masking tape.

Certain brackets can be removed to gain more room at the front end of the cage. Relays, buzzers and switches can often be moved after the cage is installed. It is possible to save significant weight in the car by removing unused seat, speaker, and convenience attachment points. Brackets that hold important controls, such as the brakes and steering system, should be left in place. There is a possibility that after the cage is installed, it will be very difficult or impossible to remove these items. Plan ahead.

The first piece of the cage to be made should be the main hoop. All cage elements will be tied to the main hoop; it is also the longest continuous tube. The main hoop should follow the contours of the car’s B-pillar and roof. This often means the hoop should have at least 5 bends, one slight bend for the peak of the roof, then 2 for each side’s shoulder. The bends on each side should be symmetrical, but the length can vary, as they can be shortened to meet the feet.

Now that the main hoop is bent, bend the door hoops, one for each side. The lengths of each end don’t matter at first. Again, start at the floor and make marks for each bend. After the tube is bent, the floor length can be adjusted and a third, A-pillar follower bend can be used. This will keep the bar quite close to the slightly curved A-pillar.



Figure 15: Main hoop and A-pillar bars, sitting next to the car.

The main spider is almost complete. Now the main hoop must be put in the car and held near or at its final resting place. Do this using welders angle magnets attached to the roof. Measure and notch the A-pillar bars, and tack-weld them into place.

Connect the two A-pillar bars with a windshield header bar. Two shallow bends about 5 inches from each end, and a bend at the middle to follow arc of the windshield, will keep the bar tight to the shell and out of view from inside. When this bar is notched, tack it into place.

The diagonal roof bar is the last one to have any bends in it. Since the main hoop sits at the top of the roof, and the windshield bar is slightly lower, this tube will need at least one downward bend near the windshield. Remember that gussets will take up space on the top edge of the windshield bar, so the diagonal will not meet at the front corner, but 4 inches inward.



Figure 16: The junction of the diagonal roof bar, the windshield bar and the driver’s side A-pillar bar. Note the lower gusset already in place, and the slight bends in the ends of the bars. Clearance is left for the upper taco gusset.

After the roof diagonal is bent, it is necessary to weld the first junction. The diagonal will cover the passenger a-pillar and main hoop joint. Carefully weld this after making sure that the cage is symmetrical. When this is done, the diagonal can be tacked into place.

Begin welding small parts of each junction in the spider. Don’t concentrate on the inside edges, as built up heat will tweak the cage. Move around and weld the outsides as well. Once the integrity of the spider is certain, it can be shifted to complete the perimeter of the welds.

Fabricate the 4 gussets next. Make a template for each out of card stock. They will look like elongated diamonds, with opposite corners cut off. After marking the inside and outside of the bend, transfer the template onto 0.120 inch thick sheet metal. Take a 2two foot length of the scrap tubing and tack the center of the gusset on one side. Clamp the other side, if it is tacked, the heat will pop the weld when the material is heated. With one side of the gusset in a vice, heat the metal to its plastic deformation temperature. When the gusset turns red, use an appropriate tool to bend the metal 180degrees around the 1.5in tubing.

Some adjustments will be necessary for the gusset to fit properly. Grinding down a bit of the weld under the gusset will give a good fit. A clamp can be used to close up some small gaps, as long as it remains in place until the weld cools.

The gussets should be carefully welded in a similar manner as the spider. After the being tacked, proceed slowly, and only shifting the cage after integrity and symmetry is attained.

If possible, put in seats now. These allow the ideal door bars to be made, and harness mounts to be studied and incorporated. Lacking seats at this point can lead to poor door bar positioning. Bars may interfere with seat position, or present a hazard in a crash by breaking a occupant’s rib instead of his seat. Harness bars must be carefully positioned, so that proper angles are maintained between the person’s shoulders and belts.

Complete the door bars. These simple, straight tubes only require notching, tacking and welding. Space them above the sill so that the undersides can be welded. If necessary they can be made longer to wedge the cage back to an original size. Shifting the cage will help welding the outsides at the main hoop and a pillar. The acute angles where door bars intersect must also be welded all the way around, with a large stitch.



Figure 17: Door bar notching and fitting.

Now the cage feet should be fabricated. Shift the cage into its final resting place. Make marks on the shell where the cage rests. Another template should be made out of cardboard, to visualize the bends, and ensure a tight, weld-able fit. Use heat to bend the 1/8in sheet metal. Build the boxes, and trim legs as necessary. Try to arrange a welding order that allows the outside leg edge to be welded into place. Some of the tightest welding will be in this section, the main spider will not move from now on so make sure it is put in straight.



Figure 18: Bending and fabricating front cage feet. Note the A-pillar bar, squeezed between the sills, to rest on the floor. Also the single bend in the plate to make the box.

The rear bars are simple and straight. Notch the upper hoop attachment carefully, as the bars intersect at a corner and an angle. Weld carefully, the top edge will be difficult, as the cage can no longer be shifted and clearance is tight. Rules state that they must be between the two a-pillar attachments.

Rear diagonals involve more careful notching. They should fit in the plane of the rear bars.

Harness mounts fit conveniently between the rear bars and rear diagonals. This allows for access to the area behind the seats and an additional tubing/weight savings.

Suspension tie bars connect rear shock towers to the base of the main hoop, which is also the rear control arm pickup point. The wheel wells were dented slightly with a hammer, to allow the bar to end closer to junctions. These are the last bars welded to the rear of the main hoop.

Main hoop V-brace, see figure 4. This was done as a two bar piece with a plate style foot on top of the transmission tunnel.

Tabs should be made to tie the cage into the shell at more points. A simple way is to order tabs that are pre-shaped and bent. A cheap way is to make twice as many by using a 1.25 in X .125 in steel sheet, cut to 1.5 in lengths. These are welded along the A and B-pillar, roof, main hoop and windshield.



Figure 19: Welded tabs and money-weld, as looking downward upon the A-pillar of the passenger side.

The money-weld is completed where the cage tubing actually touches the shell, showing off the fabricators skill. In this project, the money welds are along the top and bottom of the A-pillar.

Finishing

To protect the cage from rust, it should be painted. A spray primer and paint was all that was necessary after sharpie marks were removed with acetone. The rear of the cage is white, the theory being that cracks will be more visible on a white background. The front of the cage is a semi gloss black that will reflect very little light at the driver and collect less dust than a flat black.

Interior pieces can now be added. Re-installing the dashboard requires cutting notches on the windshield edge of the dash. If a close fit is not desired, time can be saved by cutting ample clearance, then covering gaps with a dash mat.

A windshield must be installed. With a cage installed, the windshield becomes less of a stressed member. Therefore, the highest strength adhesives are not necessary. The ability to change out a damaged windshield during an event is perhaps more relevant then strength.



Figure 20: The finished cage and a young driver.

conclusion

Producing a roll cage is a large project. Many elements of the design were dictated by the rules; other elements were dictated by interactions with other cages.

The fabrication of the cage is important to its final integrity. Skills involved in the fabrication improved with practice. Welding confidence improved greatly, as did notching skill. The fit and finish are above personal expectations and standards. I would let my mother crash this car.

This roll cage meets the 2005 Rally America rules. The Toyota is now ready to compete in Rally America events.



Acknowledgments

Noah Silverman, for fabrication assistance.

Derick and Pete, from Cascade Autosport, for knowledge and tube bending.

John Vanlandingham, various anecdotes.

References

1. Rally America Performance Rally Rules, 2005 edition. Published Nov. 2004, from www.rally-america.com

2. Tube Miter program available at http://www.ihpva.org/people/tstrike/building/tubemit.htm



CONTACT

Adam Crane,

---

"I put the hurt on dirt" - adam crane
http://CraneRallyCrew.com
corolla gt-s "Patches"
Op: S.S.