Cage Design

[Paul Bird]

Jon is correct. The reason for a horizontal 'X' in the back is to improve side impact protection of the driver. Unfortunately this would does little for flexural rigidity. If the main goal is stiffening and rollover protection then run the 'X' high and low from the rear shock towers to the main hoop.

I'm more in the 'cage failed' camp, and I don't think the scenario is as unlikely as we would like to believe. Landing on top of a Kwall, a secondary impact with a bumper after getting launched - I think similar loading could occur without a tree needing to be involved.

Ahhhh, this is refreshing.

That dirt car is real piece of work - no diagonals at all in backstay plane or main hoop! The collapse at the LF corner of the halo makes me think it's really thin wall tubing.

The real sad thing about this is that this rollover was on dirt at 50 mph. From what I can gather from other pics the tubing is actually pretty heavy, at least .120". BTW, this should be a case study for how not to build a cage. I am sure the driver thought that with the 5 door bars that he was safe as he can be.

[Boo Boo Foo]

Blainefab? May I take this opportunity to welcome you to the Corner Carvers messageboard. I'm really enjoying your posts and I suspect your presence here will be appreciated by many others for a long time too.

As you might be aware, we Australians have a V8 Supercar series which still use road going shells as the basis of a 5.0 litre pushrod formula. Cage design, engine build-ups, and suspension builds are tightly kept secrets amongst the teams and it's a very, VERY competitive series with lots of teams doing some very inventive design work.

I've managed to find a photo which, to the best of my internet searching, is the only photo I can find of the latest '08 FG01 Ford GTP car in it's build up phase without panels and doors. Sadly, it's not a large shot but it's very interesting as to how the latest race is car is getting built up, cage wise...

[LMan]

Originally Posted by Paul Bird

It doesn't stop Zee Germans:

Nothing stops zee Germans!



well, except the Russians.

[gt40]

Originally Posted by LMan

Nothing stops zee Germans!



well, except the Russians.

And the English Channel.

[79coupe]

Originally Posted by gt40

And the English Channel.

And the French 'Maginot Line' . Oh wait, nevermind.

[swhiteh3]

Originally Posted by Paul Bird

Blain,

Nicely executed cage.

Suggestion:
In your cage the upper halo section is pre-bent to fold the upper section of the windshield down onto the driver in the event of an impact at the top of the windshield (which is VERY commond in any rollover). With that, I think you are making a big mistake not putting in a secondary 'A' pillar support as I like to call them. If you look closely at the WRC car above you can see this support (it is black). This support is now mandated by the FIA and for good reason.

It is easier to show so see below.

An alternative to this would be to run the cage member from the top of the windshield down to the 'x' in the door bars. This would make for very tight entry/exit though as the center of your x is midway the door opening.

The FIA-regulated diagonal that Paul drew in is a really nice addition to roll-over protection. But I'll put in a few thoughts... In a car where the driver's head is farther forward, closer to the high node in the front downtube, then this should be nearly required.

However, in many cars (including my new cage) the main hoop and driver are so far back that the collapse of that forward node would not be a major issue for me. In my case, I decided to try to move the inevitable failure point forward but adding a diagonal brace from the mid-point of the main hoop, diagonally up and forward, to pickup the A-pillar just in front of my head (in side view). This also has much more benefit in torsional rigidity (where the FIA bar contributes very little).

Blaine's cage is very nicely fabricated, but I have serious concerns about the bends in all the top bars. I understand that there are package limitations, but structurally it's a MAJOR compromise. If that were my car, I would add BOTH the FIA rollover bars and the diagonals above. I'd probably run the FIA bar to the forward most "peak" in the top side tube, and the diagonal above to the rearward "peak" in the top side tube.

[DavidHarsay]

Originally Posted by swhiteh3

...Blaine's cage is very nicely fabricated, but I have serious concerns about the bends in all the top bars. I understand that there are package limitations, but structurally it's a MAJOR compromise...

Considering that the bends form part of a triangulated section, this seems to me to be an ingenious and appropriate way of significantly limiting the structural compromise and also addressing the packaging constraints.

[blainefab]

Originally Posted by Boo Boo Foo

Blainefab? May I take this opportunity to welcome you to the Corner Carvers messageboard.

...Cage design, engine build-ups, and suspension builds are tightly kept secrets .

Thanks for the welcome, Ive. I hope to contribute a fraction of what I've gathered from lurking here.

Secrets, Ya - Pro cars rarely get photographed with the doors open. Thanks for the pic, there are some familiar elements in that cage.

[Tom Beverly]

Scott wrote: However, in many cars (including my new cage) the main hoop and driver are so far back that the collapse of that forward node would not be a major issue for me.

The only obvious exception to this would be getting out of the car quickly when it is on the roof with the window section collapsed (and of course fuel will inevitably be leaking all over the place when this occurs).

[blainefab]

Originally Posted by swhiteh3

However, in many cars (including my new cage) the main hoop and driver are so far back that the collapse of that forward node would not be a major issue for me. In my case, I decided to try to move the inevitable failure point forward but adding a diagonal brace from the mid-point of the main hoop, diagonally up and forward, to pickup the A-pillar just in front of my head (in side view). This also has much more benefit in torsional rigidity (where the FIA bar contributes very little).

Scott - Agreed that a rear set driver position, in a sedan, makes for a cage much less compromised for access. I'm on a current build that has a vertical supplemental Apillar, and the main hoop to halo side bar diagonal, altho on the GM Fbody the greenhouse narrows at the top so limits how forward on the halo that diagonal can go without interfering with the seat - I like to leave about an inch clear from the seat halo (Racetech HR) and steel.

I would add BOTH the FIA rollover bars and the diagonals above. I'd probably run the FIA bar to the forward most "peak" in the top side tube, and the diagonal above to the rearward "peak" in the top side tube.

On the Miata the short, narrow greenhouse really makes bracing to the points you suggest conflict with providing reasonable access, and clearing the seat HR section. Are there other ways to skin this?:

-Tying the Apillar tube to the full length of the windshield frame should increase the bend resistance, since the cross section has been increased.

-Gussetting the inside radius of the halo bends - would this mitigate the bend, or just move the weak spot to the end of the gusset?

-A combo of the 2, ie pull the A pillar tube a couple inches away from the windshield frame, and then box it to the frame. I've seen this on some DTM cars.

[blainefab]

Yes, this one has braced the roof, but compromised the side impact protection for the driver. It looks like a corner of a car could get thru to the seat, and the drivers lower torso. I would not consider that acceptable in a RR car.

[b_tone]

Originally Posted by swhiteh3

The FIA-regulated diagonal that Paul drew in is a really nice addition to roll-over protection. But I'll put in a few thoughts... In a car where the driver's head is farther forward, closer to the high node in the front downtube, then this should be nearly required.

However, in many cars (including my new cage) the main hoop and driver are so far back that the collapse of that forward node would not be a major issue for me. In my case, I decided to try to move the inevitable failure point forward but adding a diagonal brace from the mid-point of the main hoop, diagonally up and forward, to pickup the A-pillar just in front of my head (in side view). This also has much more benefit in torsional rigidity (where the FIA bar contributes very little).

Blaine's cage is very nicely fabricated, but I have serious concerns about the bends in all the top bars. I understand that there are package limitations, but structurally it's a MAJOR compromise. If that were my car, I would add BOTH the FIA rollover bars and the diagonals above. I'd probably run the FIA bar to the forward most "peak" in the top side tube, and the diagonal above to the rearward "peak" in the top side tube.

can you show a picture of the portion in bold, I am having a tough time following you.

[SHOboy]

Originally Posted by swhiteh3

However, in many cars (including my new cage) the main hoop and driver are so far back that the collapse of that forward node would not be a major issue for me. In my case, I decided to try to move the inevitable failure point forward but adding a diagonal brace from the mid-point of the main hoop, diagonally up and forward, to pickup the A-pillar just in front of my head (in side view). This also has much more benefit in torsional rigidity (where the FIA bar contributes very little).

What you described has the added advantage of minimizing egress restriction. The FIA bar will be in the way to some degree in most cases, I think.

[4DoorBoss351]

Originally Posted by SHOboy

What you described has the added advantage of minimizing egress restriction. The FIA bar will be in the way to some degree in most cases, I think.

Having clambered in and out of the passengers side of the 3 year earlier sister of the car Ive posted when i scored a couple of laps in the pasengers side with Craig Lowndes, the FIA bar is a bit of a pain, but it is an extra "monkey bar" to swing off of for getting in and out - its not significantly harder than my car with the wheel off. For me the main issue was the seat was mounted ontop of a floor diagonal and the top of my helmet was above the halo bars...

[swhiteh3]

Originally Posted by Boo Boo Foo

I've managed to find a photo which, to the best of my internet searching, is the only photo I can find of the latest '08 FG01 Ford GTP car in it's build up phase without panels and doors. Sadly, it's not a large shot but it's very interesting as to how the latest race is car is getting built up, cage wise...

Great photo. Too bad it's not hi-res - there is a lot of info there. Any other similar pictures, even of older cars?

BTW, it seems that picture was taken about the same time this article was written...

Originally Posted by b_tone

can you show a picture of the portion in bold, I am having a tough time following you.

I'm on the road and don't have access to pictures and stuff. It's similar to the diagonal bar in the upper left corner of the pic that Blainefabrication posted (orange car). But take that bar and move it out, so the lower end reaches all the way to the node below it, and the upper end reaches forward much further. Does that make any sense?

[b_tone]

Originally Posted by swhiteh3

Great photo. Too bad it's not hi-res - there is a lot of info there. Any other similar pictures, even of older cars?

BTW, it seems that picture was taken about the same time this article was written...

I'm on the road and don't have access to pictures and stuff. It's similar to the diagonal bar in the upper left corner of the pic that Blainefabrication posted (orange car). But take that bar and move it out, so the lower end reaches all the way to the node below it, and the upper end reaches forward much further. Does that make any sense?

I'm going to need a picture, my brain is not working well.

[claykos]

I believe Scott means something like this....

(the bar on the right in red) but maybe with the mount to the halo further forward.

[blainefab]

I mocked up what I'm thinking for a bend reinforcement gusset - It extends 1.5 tube diameter fore and aft of the ends of the bend, and broadens the section width by 50% on the side pointing at the driver. I'd use pad stock (.100" HR), fully perimeter welded, no holes.

With the gusset under each of the halo bends, and a compromise position for the diagonals that Paul and Scott suggested, there would be minimal impact on access, about 8# of steel.

pics below, revisions in green. Comments?

[Paul Bird]

Originally Posted by swhiteh3

This also has much more benefit in torsional rigidity (where the FIA bar contributes very little).

While I agree with most of what you said, this is very, VERY wrong. Model the secondary A pillar support and you will see. This cage member is a major win, win.

Alan,

I like your revisions above. The slightly shifted secondary A pillar is similar to what I did on my car. To me, the key to shifting this forward is to line it up with the gusset on the windshield opening. Below is a picture of my car in which you can see that the secondary A support runs right into the tubular gusset which then continues into the diagonal of the halo.

[DavidHarsay]

Paul,

That's such a nice clean cage... what are your thoughts about eliminating the main hoop horizontal cross bar or harness bar?

Dave

[Paul Bird]

That's such a nice clean cage... what are your thoughts about eliminating the main hoop horizontal cross bar or harness bar?

Clearly I don't have a problem with it. The harness is mounted to the gusset on the rear diagonals. BTW, this picture is from a long time ago and things have changed somewhat since. The oil tank is now adjacent to the fuel cell and cage members are going in from the main hoop at the upper door bar intersection back to the rear shock tower junctions.

The is a picture of the cage from the 2005 BMW M3 GTR. Here you can see how it is going to look. Note that they went one step further and eliminated the main hoop diagonal all together.

[MTNMAN]

I thought the FR500C cage design was kinda interesting.

[Paul Bird]

The door bars are too low and leave an opening in the cage big enough to almost stick a whole bumper through. Plus, if you asked me to make my door bars with two parallel members so that it added as little rigidity as possible, this is what I would come up with.

The A pillar cage member should follow the stock A pillar as to not create a visual obstruction that is twice that of the stock A pillar - which is saying something.

"Petty Bar" - if the purpose of this is additional rollover protection (which I don't think is needed as the main hoop is well supported) then fine. The problem is that the additional stiffness that this would provide is almost completely negated by the fact that it intersects the main hoop mid-span.

It is hard to tell what is going on behind the dash but I am pretty sure that there is a big unsported bend at the upper dash on the A pillar cage member. Through the firewall is a mystery as well. More pictures are needed.

My grade on this cage is a C. They get this because the junctions appear to be well executed and the horizontal and vertical cross sections are maximized.

[SHOboy]

Doesn't look like the FR500 cage would pass NASA tech since the only rear support for the main hoop is the diagonals? i.e., no downbars going roughly straight to rear support.

[MTNMAN]

[b_tone]

Originally Posted by SHOboy

Doesn't look like the FR500 cage would pass NASA tech since the only rear support for the main hoop is the diagonals? i.e., no downbars going roughly straight to rear support.

This is incorrect. Since the car has been homologated for use in Grand Am the car is 100% legal for NASA use. That's not to say that having the point of failure be a single point for the rear main hoop supports is a good idea......

[Bossbill]

Is there a preferred die radius for cage work?
For instance the dies available for 1.5" tubing from one vendor are 5", 6" and 7".

Or this another one of those, "well, it depends"? If so, help me out and give me some guidelines.

[blainefab]

Originally Posted by Bossbill

Is there a preferred die radius for cage work?
For instance the dies available for 1.5" tubing from one vendor are 5", 6" and 7".

Or this another one of those, "well, it depends"? If so, help me out and give me some guidelines.

SCCA and NASA spec bend CLR at minimum 3X diameter. In practice, the longer the radius the less distortion of the tube, with thinner wall tubing distorting more than thicker. I use 1.5, 1.75, and 2" dies, all in 6.5" CLR.

If you're going to be bending .095 wall, I'd go with the 7"

[blainefab]

Originally Posted by SHOboy

Doesn't look like the FR500 cage would pass NASA tech since the only rear support for the main hoop is the diagonals? i.e., no downbars going roughly straight to rear support.

The cage would fail for that reason, but the car would likely be given a pass if it had a logbook and current yearly from another sanction.

Here, a FR500S has conventional, proper backstays, single diagonal in the backstay plane, full width harness bar, and a main hoop welded to the chassis:

[blainefab]

Originally Posted by Paul Bird

My grade on this cage is a C. They get this because the junctions appear to be well executed and the horizontal and vertical cross sections are maximized.

I'd be much tougher on it - the lack of backstays rate it an F. Fix that and the driver door bars give it a D - those S bends on each end look ready to bend right into the driver with a good side impact, and the supplemental A pillar support lands right on the springy part of the door bars.

For having the same corporate logo, it is quite a contrast with the FG01 GTP car.

[StanIROCZ]

Originally Posted by blainefab

Comments?

I wouldn’t change it.



The area to the lower left of the orange line I drew isn’t going to budge. That’s the firewall/bulkhead area where that the door is bolted to and is likely very beefy from the factory. Now he’s added the additional 1.5” tube spaced ~2” away with the plate/gusset in-between which is welded the full length which makes it roughly a 10” cross sectional beam (red/pink arrow)+ the additional support of the side impact bar which is mostly triangulated + the gusset above that.

Pretend bar “A” wasn’t there. If an impact was concentrated in the area where I drew the green arrow the windshield would bend where I drew the orange line. But the cross section at the orange line (and above) is significantly increased. I could approximate some MOA’s for before vs. after the addition of the tube, but I think you get the idea. It is now very stout since the tube is fully integrated into the windshield frame. Now add bar “A” into the picture and you can see the additional support given.

Yes the FIA bar would make the cage stronger, but in this case given significance of the reinforced beam from the floor to the top of the windshield bar I see it as diminishing returns.

It’s a accident probability game. You could find compromises in any cage on any car. If you didn’t find compromises it would probably be too heavy to ever win a race. The key in my mind making reasonable compromises that gain the highest probably of survival with the least amount of weight increase. I like how Blaine did this cage.

Originally Posted by swhiteh3

Blaine's cage is very nicely fabricated, but I have serious concerns about the bends in all the top bars. I understand that there are package limitations, but structurally it's a MAJOR compromise.

More important than strength is energy absorption, F*d. A bend creates a thinner wall thickness on the outside radius therefore less strength, but the benefit is increased distance before contacting the drivers head. A small reduction in strength/stiffness but a larger increase in distance increases the amount of kinetic energy absorbed in an impact. The red bar "A" that I drew would have been the alternative to adding the bends. It’s even closer to the drivers head than it appears in this view since it would move in 2 planes.

[Preston]

Maybe a non sequiter but those DTM cars also have large styrofoam filled carbon fiber crash pods on the each side. Wonder how that affects your cage side impact design?

[blainefab]

Originally Posted by Preston

Maybe a non sequiter but those DTM cars also have large styrofoam filled carbon fiber crash pods on the each side. Wonder how that affects your cage side impact design?

I expect there is more to the side impact system on the DTM car than is revealed by that pic, but a similar function on the Pratt & Miller WC cars, and the COT, have substantial cage backing up the foam pods. A least one of the P&M cars have gone to box section tube in the door bars, I assume for better bend resistance.

I'm intrigued by the benefits of the IMPAXX energy absorbers, but I think it would need some platform specific engineering to make the jump to club race cars.

[swhiteh3]

Originally Posted by Preston

Maybe a non sequiter but those DTM cars also have large styrofoam filled carbon fiber crash pods on the each side. Wonder how that affects your cage side impact design?

The DTM cars are so far from anything we're working with that it's nearly impossible to look at them for structural ideas or direction. WRC or Australian V8 Supercars are more more similiar.

[swhiteh3]

Originally Posted by claykos

I believe Scott means something like this....

(the bar on the right in red) but maybe with the mount to the halo further forward.

Absolutely Correct.

[swhiteh3]

Originally Posted by Paul Bird

While I agree with most of what you said, this is very, VERY wrong. Model the secondary A pillar support and you will see. This cage member is a major win, win.

Paul- I say this based on two FEA studies done by professional teams in FIA series' which studied just that when the bar was mandated. If you have properly dealt with the door opening already, then there is nearly no advantage to the FIA bar in torsional or bending rigidity.

On Edit: BTW Paul, with that said I really like the way you handled it in your car. It seems to replace the gusset you'd typically find at the front of the window opening, and tie nicely into the door X bars. I like it - much more efficient than a typical "have a full cage then add the FIA bar as a crude afterthought" approach.... Have any other pics of the car recently?

[b_tone]

Here's an example of what can be considered a rather "normal" cage being put to the test and it holds up well enough. There are no detailed shots of the cage, but its a T2 car so the cage is not exotic.

You can see there are no FIA bars, standard NASCAR type door bars, the A-Pillars are not tied in via welding to the cage etc.

Just food for thought as this thread can be viewed in the same light as the crazy guy on the corner screaming for everyone to repent because the world is coming to an end.

[Tom Beverly]

WOW, now that is an impressive set of photographs... Is that the same turn the DB9 Lemans car flew 142'?
I think they need a larger catch fence.
Glad to read he walked away.

[Bossbill]

Here is the 142 footer:
http://www.youtube.com/watch?v=hl2UUunlI2Q

[Paul Bird]

Paul- I say this based on two FEA studies done by professional teams in FIA series' which studied just that when the bar was mandated. If you have properly dealt with the door opening already, then there is nearly no advantage to the FIA bar in torsional or bending rigidity.

I ran this through Grape and got a significant benefit. Also there was an article in Racecar Engineering four or five years ago on a British Touring car build that noted that their torsional figures jumped 20% from the addition of the secondary A supports alone. BTW, this car was quoted to have 45,000 ft.lbs/deg.

ere's an example of what can be considered a rather "normal" cage being put to the test and it holds up well enough. There are no detailed shots of the cage, but its a T2 car so the cage is not exotic.

You can see there are no FIA bars, standard NASCAR type door bars, the A-Pillars are not tied in via welding to the cage etc.

Just food for thought as this thread can be viewed in the same light as the crazy guy on the corner screaming for everyone to repent because the world is coming to an end.

Wow! That was a hell of an off. The safety equipment worked very well. Of note is that the stock crumple zones did their work well and this is why I don't advocate running the cage beyond the suspension pickup points on a production based car.

With that, the only impact to the roof was at the very end of the wreck and was of very little magnitiude. There was no door impact whatsoever.