Project Professional Awesome Time Attack Evo: Part 1 - Roll Cage and Chassis in Detail

by Daniel O'Donnell

Hi! For those that don’t know me, I drive the Professional Awesome Evo from time to time. That being said, I am no engineer, I don’t weld, I don’t fabricate, but when it comes to pushing the broom around the shop and holding things while the guys tack important bits of metal in place, I’m a champ. This should be kept in mind when reading this article. I may not be a complete dip sh!t, but I don’t know everything there is to know on this topic. Please don’t hesitate to comment and add to this conversation, let me know where I’m right and where I should get to studying!

Since we’ve passed the disclaimer I should be honest with you, I was supposed to update you along the way of the build of our V2.0 Time Attack Evo. I wrote the initial article, but after that I didn’t complete my task and I blame everyone but myself. I’ve decided, after gentle prompting from Mike Kojima (ie, “Where the hell is the damned story?”), to finish what I started writing months and months ago and give you all the details I can possibly give about the car. Without further ado, here’s part 1.


This was our starting point. A recovered, stolen Evo that once had a life as a magazine car. Luckily for us, and unluckily for the previous owner, thieves stripped the majority of the interior in advance.
Well documented in previous Moto IQ articles (here and here), we started the roll cage process by stripping what was left of the interior and then using dry ice to remove the sound deadening.
This involved using some tricky techniques to get vertical pieces of sound deadening off. We put unpaid intern Zack on the task as he's the most flexible of the bunch. Read into that what you will.
We also removed all the glass, doors and roof from the car to make cage install easier. This allows for tighter tolerances between the cage and the chassis roof. This can translate to more clearance between the driver and those bars, which enhances safety.
Page 1 of 7 Next Page
Bookmark and Share
Sunday, February 08, 2015 11:56 PM
Thanks for the article, please keep them coming :D
Name User
Name Userlink
Monday, February 09, 2015 12:36 AM
Thanks, totally stopped my Sunday night to read through this in a frenzy.

Now that I understand the magnitude of this job, I ended up making this thread on evom: http://www.evolutionm.net/forums/evo-general/707686-high-mileage-chassis-rigidity-options-without-rollcage.html

Does the CT9A get all wobbly with high mileage abuse? What would you do for a long term canyon car? I'm talking 200,000+ miles.
Monday, February 09, 2015 6:26 AM
My XP is on the S13 chassis, but i guess it applies to all.

1- Foam does jack, some basics maths around it proves it.

2 -adjustable braces usually do not reinforce much either. Most of them are badly design, the bolting is usually no good. Welded braces, much better.
I used to put every bolt in brace i could find, but most did nothing. The "nismo power brace", which consist of new, boxed front caster arm supports braced together had a noticeable effect, but that was the only one. I run semi slick tyres, so maybe bolt in braces have more effects on slicks ... but i don't think so.

3 - roll cage. It works. I only got a rear half cage, but that really stiffened the car, in a way no amount of braces ever did.
However, as i felt the rear of the car go solid, it did not increase front rigidity and i can feel how the front is flexible.

I did not stitch weld it for the same reason i went half, bolted cage : did not want to strip the whole car, so i cannot comment on it. Considering how the unibody is welded, my guess is it improves rigidity too.

Monday, February 09, 2015 7:12 AM
I think FIA-spec cages require an additional, more vertical tube to support the A-pillar in the event of a roll over. Due to the long span between the steeply raked A-pillar segment of your cage and the main hoop, there is nothing to support or brace this relatively horizontal section. Since the driver is behind the car's B-pillar, maybe it's not as important if the roof crushes down at the A-pillar to the B-pillar?
Monday, February 09, 2015 7:16 AM
If you notice on page 3 we had the A-pillar support in the design, but there were a few concerns we had so it was decided against for the short term. Don't be surprised if you see it added in soon though!
Monday, February 09, 2015 8:34 AM
I saw an S14 on Craigslist out here that broke every rule of rollcage design. Tubes were massively oversized, there were far too many bends in the main hoops (all done to avoid the dash or the front occupants, which is really dumbe since you could just move the main hoop back), none of the downtubes had floor supports, and there were unsupported nodes everywhere. The ad is down now so it seems some poor sap bought this car. I always shed a tear when I see a nice car like that ruined by the previous owner's mid-directed fabrication.

The door bracing through the B-pillar is very clever. I hope more car builders borrow that idea.
Monday, February 09, 2015 8:51 AM
Great build, look forward to more on it.

Anybody ever actually acid dipped a car on here?

When I was looking into it all I found were some people saying to not do it as acid ends up in the seams and comes out days/months/years later killing the paint and causing crevice corrosion. Seemed like most recommended media blasting over acid dipping that had done it.
Monday, February 09, 2015 9:00 AM
We talked about alkaline dipping and then e-coating the car (which is another dipping process). Alkaline is apparently easier on the metal than acid and then with an e-coat, everything is sealed and should really reduce any chances of rust. See here for more info: http://www.autorestorationdepot.com/about.html
Monday, February 09, 2015 10:30 AM
"We use the DOM material in our cage design."

As far as I know, DOM is a manufacturing process and not a material and you can have 4130 (chromoly) that is considered DOM. As a correction, what you mean to compare is mild steel to 4130 (chromoly).

Does anyone have a good reference/information on steel tubing? There are a lot of acronyms and ambiguous claims to sort through on the web.
A. F.
A. F.link
Monday, February 09, 2015 11:48 AM
The 'guesset trade study' on page 2 is interesting. Adding more material in sheer will definitely increase bending stiffness, but
1) the lower plots show the same deflection
2) the car wasn't built like the FEA test

Why not just get the cage fitment tight and weld the bar edge to the a-pillar (z=0mm gusset)?

There is a lot to be said for a less fabrication intensive design (time is money), and doing this while making a customer happy is made easier with FEA graphs. Most of an engineer's job is selling their ideas to people that don't understand engineering.

I strongly support dipping cars after going a heavy sandblast route a few times. Finding someone that can do it on modern paints (BMW paint is apparently stronger than eco-safe strippers) is tough. Definitely cut any brackets and body (like rear quarters) before getting to the blaster/stripper though!
Monday, February 09, 2015 12:07 PM
The large note next to each FEA plot is a typo, deflection was not the same in each model, the note on top is accurate, 84% reduction in deflection with the gusset in the representative trade study is what we found and is what should have been highlighted. I fat fingered the text box back then.

While the cage was not built exactly as the sketch indicates, it is similar, the point was more to show how the gusset will reduce deflection and reduce stress in the tube by creating an I-Beam like cross section out of the A-Pillar structure.
Monday, February 09, 2015 12:36 PM
Looking at it, it seems pretty clear whats happening. More deflection is taking place in the outer A-Pillar and less is transfering into the cage material. That small link between the two is going to buckle before it can transfer a substaintial amount of load into the cage material.

It would be interesting to see total energy absorbed using a fixed deflection of the cage material. I wouldn't be suprised if it was the gusseted piece still won out, but mostly because it looks like there would just be more energy absorbed by the A-Pillar before it reached the allowed deflection. There is more material to deform on the gusset too, which is a dead give away it will absorb more energy (and weigh more).

Cool stuff either way.
Monday, February 09, 2015 2:17 PM
I am wondering though, obviously you did what you could to prevent dead nodes. And while the strut towers to the dash bar are not "T-angle" 90' junctions. They do still present something of a concern, to me at least. I suppose with a big-stiff shock-tower brace the chances would be reduced, encouraging side-to-side load transfer, but it still looks like in a front corner impact you'd be transferring forces directly into the dash bar, which stands the greatest risk of penetrating the occupant area.

Regardless very cool project, sorry you went through the pains with seam sealer and the unneeded brackets after the fact.
Monday, February 09, 2015 2:35 PM
do you need the car on a frame jig to do the stitch/seam welding?

I thought all that constant heat would warp the metal?
Monday, February 09, 2015 3:17 PM
"We wonder how our 206lb cage compares with the one in the Cyber Evo."

I'll give you a hint: it's heavier.

This is how the Cyber Evo people probably view your logic:

"Oh bother, I've just crashed. Hmm... What can I do to make further track-driving a little safer for myself?

Hone my driving skills to avoid loss of control in the first place? Nahhh, 'mediocre' is good enough.

Be a little more careful and choose 9/10 over 11/10 in the future? HAHAHAHA!

F*ck it, add some weight to the roll cage."

To which their response would likely be: "Ehh nani!?"
Monday, February 09, 2015 5:11 PM
It's really nice to see a proper cage being built and documented, and the hows and whys. There's way too many brightly-painted examples of poor cage design and poor attention to detail all over the blog world, and the mistakes are so easy to spot.

I was going to make a joke (from the opening photo) about car theft getting really bad, and then I read the captions and realized it really was a theft victim. That's nothing to laugh at.
Tuesday, February 10, 2015 8:25 AM
When I look at this it makes me want to go and start cutting more and more off my car haha, great article and design.
Thursday, February 12, 2015 7:16 AM
dipping the chassis? does this mean acid dipping it to remove all of the paint and the sealant?
Thursday, February 12, 2015 1:18 PM
Yes, but in our case we looked into alkaline dipping.
Friday, February 13, 2015 1:34 PM
The gusset design is stronger in shear because deformation in shear is dependent on area moment of inertia. basically the shape of the cross section and how de-centralized the majority of it is in the plane of the applied force. The gusset spaces the areas of the a-pillar and the tube out further as the article says, just like an I-beam.

The devil's advocate would ask how much to space it. you could have a 2 foot wide gusset plate and that FEA analysis would show almost no deflection. It's also true that they are only showing deflection in the most ideal circumstance possible, force applied on the axis of the gusset. The cross section makes the A pillar much stronger in that particular axis but perhaps more prone to twisting. It's for this reason that I-Beams are used only to support vertical loads, their shape has been optimized for that purpose only.

As with all engineering it's a compromise so things are only really right or wrong to degrees. The builder's made informed decisions to build the car according to their priorities so I wouldn't critique it, but people shouldn't think that one way or another is the gospel, everything is a choice.
Friday, February 13, 2015 2:00 PM
Well said Fuergrissa.
Friday, February 13, 2015 11:53 PM
Pat, once again fantastic job on the cage! I remember when you were designing it you posted some pictures and I was baffled by the results, and it was mentioned again here.

You selected 1.5x.120 tube over the 1.75x.095 tube. This article says it was 5% lighter going with the 1.5, but the 1.75 is actually 8% lighter. (Cross sectional area is smaller). Additionally, the 1.75 also has a 30% higher I value so it's significantly stiffer. I've checked my math a dozen times on this because I'm doing it by hand and you were using FEA/cad to give you weights and strength. Am I mistaken?

(Ps, this is kevin at evo dynamics/my shop assist)
Thursday, February 19, 2015 8:56 AM
@ MDR: While you are correct, in the motorsports world everyone understands what you mean when you say DOM vice 4130 or chromoly, and I'm sure you did as well. If they had just said mild steel, that could be ERW or DOM, with very different material properties. Saying mild steel DOM every time (or 1020/1018 DOM) would get a little tedious. :)

That being said, to expand a little on the article for DOM vs. 4130 strengths, etc, properly TIG-welded (or gas welded) 4130 using ER70S-2 or ER80S-D2 filler material (vice 4130 filler) will have more tensile and yield strength than DOM, but will also (surprisingly) be more ductile than 1018/1020 DOM. Welding normalized 4130 tubing under .125" and using ER70S-2 or ER80S-D2 filler also will not require heat treating. To reduce hydrogen embrittlement with 4130, however, it is always recommended to preheat the tubing to 300*F, especially if it's cold out.

If you use 4130 filler without stress-relieving the workpiece, then yes, you're more likely to have problems with cracking.

I can't speak to folding vs. bending, however, haven't seen anything on that.
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:

© 2018 MotoIQ.com