Jump to content


Photo

Aluminium roll cages coming back?


  • Please log in to reply
30 replies to this topic

#1 MatsNorway

MatsNorway
  • Member

  • 2,822 posts
  • Joined: December 09

Posted 27 August 2012 - 19:55

One question. Why is it not legal anymore.. Does it not have better strength weight ratio?

I know Porsche used it. I believe it was in the 1970s with the 930 and/or 935s Porsches.

And it looks likey they are doing a comeback. I believe it has been made legal or discussed in a modern race series. Anyone got info on it?

Monster Tajima has what seems like a alu roll cage in hes EV pikes peak car.

Thread trigger: http://mylifeatspeed...archives/16773/

Advertisement

#2 PhilG

PhilG
  • Member

  • 483 posts
  • Joined: September 06

Posted 27 August 2012 - 21:44

its not a good material for this application, it yields far too easily , doesnt offer the same level of protection as steel , plus its a pain to weld without affecting its properties.

#3 kikiturbo2

kikiturbo2
  • Member

  • 869 posts
  • Joined: December 04

Posted 28 August 2012 - 08:16

porche used it as chassis material in welded spaceframe form. It doesn't have good stiffness to weight properties when used in such a structure as well as previously mentioned problems with welding. On top of that, it has no fatigue limit, which might not be a problem with high budget racing, but it is a problem nevertheless...

Having said that... I know of a company that wants to make titanium roll bars legal....

#4 Greg Locock

Greg Locock
  • Member

  • 6,353 posts
  • Joined: March 03

Posted 28 August 2012 - 11:02

A highly respected materials guy told me that the so called endurance limit for steels is actually a misinterpretation of the data, and poor test technique. I wish i could remember his explanation. The interwebs disagree, of course, since the interweb is a conservative medium.

#5 Lee Nicolle

Lee Nicolle
  • Member

  • 11,038 posts
  • Joined: July 08

Posted 28 August 2012 - 11:03

Steel cage is far stronger, cheaper and in the end not much heavier.You get more strength with smaller tubing, and smaller tubing is easier to fabricate. Chrome moly is even better and probably still cheaper than alloy.
Interesting scenario though with classic cars that went from alloy to steel. To race them as a classic they must be original but you cannot log book a car with a new alloy cage.
As for Titanium, hoo boy. Maybe ok for a huge dollar operation but even then doubtfull with the welding. So no and No

#6 kikiturbo2

kikiturbo2
  • Member

  • 869 posts
  • Joined: December 04

Posted 28 August 2012 - 14:36

As for Titanium, hoo boy. Maybe ok for a huge dollar operation but even then doubtfull with the welding. So no and No


try finding a youtube video about Lancia - Abarth WRC Delta Integrale programme... (here it is, six parts... )... lots of nice big chambers for titanium welding at that time at Abarth.. :)

As for current titanium applications in that respect.. I heard it straight from the "horses mouth" that they want to try it... however I also doubt that it is good material for this apllication. Especially because the material currenntly used is seam welded tubing..
Current high end WRC cars use 15CDV6 steel mostly.. EXPENSIVE..

#7 MatsNorway

MatsNorway
  • Member

  • 2,822 posts
  • Joined: December 09

Posted 28 August 2012 - 17:04

Steel cages looses some of it strenght in the transient areas too does it not..

gets down to half its strenght ive been told. Who explains the common usage of extruded profiles. yeye easy to do too.


To get back the strenght im guessing homogenisation would help.

http://www.scientific.net/MSF.693.264

And why not bolted connections? If planes can use it cars can use it.

You make a small part and make it go trought homogenisation. (saves you the mega oven) Then bolt the pre fabbed ends together. its all about clever design and surface area regardless.


http://www.supercars.../cars/5288.html
"Naturally, the car was completely stripped and Porsche managed to reduce weight to 900 kg. With ballast the car could then meet the required 970 kg formula"

This was the 3L turbo car...

Edited by MatsNorway, 28 August 2012 - 17:05.


#8 Sisyphus

Sisyphus
  • Member

  • 242 posts
  • Joined: May 07

Posted 28 August 2012 - 19:54

What is important for a roll cage is the ability to absorb energy in plastic deformation. You want as much elongation as possible so that the area under the stress-strain curve is maximum.

The amount of energy absorbed per unit volume is roughly the average of ultimate strength and yield strength, times the elongation.

So, for 6061T6 aluminum: (43 + 38)/2 * 0.10 = 4.0
for annealed 6061 aluminum: (18 + 8)/2 * .25 = 3.2
normalized 4130 steel: (90 + 70)/2 * 0.12 = 9.6
titanium 6-4: (134 + 126)/2 * 0.08 = 10.4

That isn't the only consideration but you can see the issue--you need a lot more aluminum to absorb the energy of a steel tube plus you have the concerns about softening due to welding and having a fatigue crack develop in a weld attaching the roll bar to the rest of the frame.

#9 PhilG

PhilG
  • Member

  • 483 posts
  • Joined: September 06

Posted 28 August 2012 - 21:04

Aluminium has been used for Anti Roll bars already , works fine.

Steel for roll cages is a debate that can go long and hard , traditionally T45 tube was the main tube used , but newer , supposedly better alloys became more readily available , that allowed the same strength , namely 15cdv6 which is a french spec material , which allowed a lighter cage to be made , with increased stiffness, but sadly the notch resistance of the material is nowhere near as good , once the tube has been compromised, so its doesnt bend just deform and collapse.

#10 Greg Locock

Greg Locock
  • Member

  • 6,353 posts
  • Joined: March 03

Posted 28 August 2012 - 21:32

If everyone is forced to use the same material everyone has roughly the same weight penalty. Allowing Ti gives another advantage to the richer teams, and at the same time probably results in a less safe design. So it's lose-lose. Same really applies to aluminium. Incidentally, yes you could make it from a good alloy and heat treat it post welding. You'll need a very big hot oven, and need to do something very clever about controlling deformation.

Bolts increase the local stresses by a factor of three, which can be designed around but it would be impossible to inspect for. It's pretty easy to check if a weld is good and continuous, far harder to see if a bolted design is well done.

#11 Magoo

Magoo
  • Member

  • 3,703 posts
  • Joined: October 10

Posted 28 August 2012 - 22:08

A highly respected materials guy told me that the so called endurance limit for steels is actually a misinterpretation of the data, and poor test technique. I wish i could remember his explanation. The interwebs disagree, of course, since the interweb is a conservative medium.


Exactly so. Steel doesn't really have an endurance limit. Given enough cycles, even the smallest load will eventually produce a fatigue failure -- though it may be at some point beyond the test regime's ability to identify it.

#12 MatsNorway

MatsNorway
  • Member

  • 2,822 posts
  • Joined: December 09

Posted 29 August 2012 - 06:49

So, for 6061T6 aluminum: (43 + 38)/2 * 0.10 = 4.0
for annealed 6061 aluminum: (18 + 8)/2 * .25 = 3.2
normalized 4130 steel: (90 + 70)/2 * 0.12 = 9.6
titanium 6-4: (134 + 126)/2 * 0.08 = 10.4


This speaks for itself. Steel is what they should use. That said 6061 etc is crap. Find a better comparison.

Edited by MatsNorway, 29 August 2012 - 06:53.


#13 desmo

desmo
  • Tech Forum Host

  • 29,391 posts
  • Joined: January 00

Posted 29 August 2012 - 15:17

6061 is not "crap", it is for many applications the most rational spec. It is cheap, corrosion resistant, weldable, reasonably strong after heat treatment and has essentially the same modulus as the most expensive Al alloys, so is a likely candidate for stiffness constrained applications. There are essentially no "crap" engineering metals or alloys, just ones that are misapplied.

#14 bigleagueslider

bigleagueslider
  • Member

  • 1,235 posts
  • Joined: March 11

Posted 30 August 2012 - 01:00

Bolts increase the local stresses by a factor of three, which can be designed around but it would be impossible to inspect for. It's pretty easy to check if a weld is good and continuous, far harder to see if a bolted design is well done.


Greg Locock,

With regards to analysis and reliability of aircraft structural designs the opposite is true. That's why you don't see fusion welding processes used in aircraft primary structures. With mechanically fastened joints, every step in the process (material traceability, heat treatment, drilling, reaming, torquing, etc.) can be tightly controlled and validated. The same is not true with fusion welding. That's why a typical weld analysis requires large knock-down factors for things like weld porosity, undercutting, lack of penetration, irregular surface finish, etc.

Regarding the other comments about fatigue or the capability to absorb energy through strain, they kind of miss the point with a roll hoop structure. The roll structures are not subject to fatigue loads, so fatigue life is not an issue. The roll structures are also not the same as impact structures, like nose cones or side pods. The roll structures should be as stiff and strong as possible. The roll structures are intended to maintain a protective envelope around the driver, and not to provide energy dissipating crumple structures.

Lastly, in reality most steel alloys actually do have an endurance limit. Of course determining where it lies on the S-N curve can be tricky. Fatigue life is a statistically derived value. Thus in order to generate reliable data for establishing an endurance limit one would need to test a statistically relevant (ie. very large) number of identical material samples under controlled conditions for something like 1x10^10 cycles. Obviously this requires massive amounts of time and money, so that's why you don't see a lot of fatigue data published for most materials. There is also the load/lifecycle profile to consider. What are the peak/mean loads and when do they occur in the life of the component? Being a statistical value, component fatigue life is usually expressed using an "Lx" value. For example, an L10 fatigue life of 1000 hours would imply that 90% of a given quantity of components would last 1000 hours or more for the given loading profile.

Regards,
slider

#15 Lee Nicolle

Lee Nicolle
  • Member

  • 11,038 posts
  • Joined: July 08

Posted 30 August 2012 - 11:07

6061 is not "crap", it is for many applications the most rational spec. It is cheap, corrosion resistant, weldable, reasonably strong after heat treatment and has essentially the same modulus as the most expensive Al alloys, so is a likely candidate for stiffness constrained applications. There are essentially no "crap" engineering metals or alloys, just ones that are misapplied.

A bit hard to heat treat a rollcage that is built inside a car!
I have seen a lot of both welded and bolt together alloy cages which doan ok job as a roll bar in a tintop. I have seen welds break though the bolted ones flex around far more.
And you cannot weld alloy to steel.
Todays quality steel cages far exceed any alloy cages for stiffness and rollover resistance. I spoke to a off road buggy builder and roll cage constructor and he laughed about alloy.
Even a rollcage built from hat is basically exhaust pipe is far better than alloy.
And allowing titanium as Greg said gives an advantage to rich teams.
The fancy TIG welded cages are generally not as strong as a plain MIG welded one.Actually a nickel bronzed one is better again but too slow to build. I have seem broken TIGed joints but never MIG or nickle bronzed ones.

#16 MatsNorway

MatsNorway
  • Member

  • 2,822 posts
  • Joined: December 09

Posted 30 August 2012 - 13:04

Nickel bronze?

Soldering?

Ive heard about TIG giving very sudden/adrupt heat sones

#17 Catalina Park

Catalina Park
  • Member

  • 6,768 posts
  • Joined: July 01

Posted 30 August 2012 - 13:35

At Eastern Creek about 12 years ago or so, I was driving the Medical chase car.
At the start of a combined tin top race a Peugeot 205 Improved Production car up the front of the field broke a driveshaft and was left sitting on the line while the whole field streamed through. A Ford Anglia Sports Sedan starting from near the back of the grid hit the Peugeot square in the bum rupturing the fuel tank and spinning the car down the track. The spinning actually forced the burning fuel away from the inside of the car and as soon as the car stopped the driver undid his belts and stepped out.
By then I was pulling up in the chase car and the paramedic went for the Anglia driver and the medic went for the Peugeot driver while I was left to oversee safety of the scene. I went over to the burning Peugeot to see if I could assist the fire marshal that was dousing the car (I told him his foot was on fire as he was leaning into the car and burning petrol was running out from underneath!) and just generally look at the cockpit to see how much force the driver had copped in the crash.

The thing I noticed was the alloy cage was destroyed. The car had been hit squarely in the back but all the mounting points of the cage has broken off above the floor, the rear supports had been pushed forwards and this has sheared the main hoop supports because the cross bracing had just pushed the whole cage forward in the car. The forward support next to the drivers legs had broken off in two places and a length of tube about six inches long had been ejected from the car during the spins. If the car had rolled after being hit the cage would or could have done more harm than good. There was no roll over protection left.

This was a good quality cage that was built in the 90s just before the ban on alloy cages was enforced in Australia.

#18 mariner

mariner
  • Member

  • 2,329 posts
  • Joined: January 07

Posted 30 August 2012 - 16:42

Some way back Racecar enginneering did an article on doing a Carbon Fibre Roll Cage ( as opposed to a carbon fibre pyramid as in F1).

I havent looked the article up but it sounded like an awful lot of work and testing for not to much gain vs steel

#19 Lee Nicolle

Lee Nicolle
  • Member

  • 11,038 posts
  • Joined: July 08

Posted 31 August 2012 - 11:06

Some way back Racecar enginneering did an article on doing a Carbon Fibre Roll Cage ( as opposed to a carbon fibre pyramid as in F1).

I havent looked the article up but it sounded like an awful lot of work and testing for not to much gain vs steel

And it is on a constant state of decline too.

Advertisement

#20 bigleagueslider

bigleagueslider
  • Member

  • 1,235 posts
  • Joined: March 11

Posted 01 September 2012 - 06:06

Some way back Racecar enginneering did an article on doing a Carbon Fibre Roll Cage ( as opposed to a carbon fibre pyramid as in F1).
I havent looked the article up but it sounded like an awful lot of work and testing for not to much gain vs steel


Composite roll structures actually perform quite well on single seat open wheel cars with composite tubs. Think about some of the recent crashes in F1 where the car has flipped. The drivers usually walk away. That wouldn't have likely happened 25 years ago when the tubs were aluminum and the roll hoops were welded alloy steel.


#21 Rasputin

Rasputin
  • Member

  • 960 posts
  • Joined: February 10

Posted 22 September 2012 - 18:37

What is important for a roll cage is the ability to absorb energy in plastic deformation. You want as much elongation as possible so that the area under the stress-strain curve is maximum.
...


I beg to differ, I believe the general purpose of the roll cage is NOT to collapse, in order to protect the driver, why I believe yield strength over density is what is important;

Aluminium 7075-T6: Yield 450 MPa, Density 2700 kg/m^3: Yield/Density: 167 000

Steel 4130: Yield 550 MPa, Density 7800 kg/m^3: Yield over Density = 71 000

Why I would go for the 7075-T6 "Aircraft Aluminium" any day of the week.




#22 Greg Locock

Greg Locock
  • Member

  • 6,353 posts
  • Joined: March 03

Posted 22 September 2012 - 22:44

I beg to differ, I believe the general purpose of the roll cage is NOT to collapse, in order to protect the driver, why I believe yield strength over density is what is important;

Aluminium 7075-T6: Yield 450 MPa, Density 2700 kg/m^3: Yield/Density: 167 000

Steel 4130: Yield 550 MPa, Density 7800 kg/m^3: Yield over Density = 71 000

Why I would go for the 7075-T6 "Aircraft Aluminium" any day of the week.


You are very fortunate that every day of the week you have access to a heat treatment oven big enough to take your roll cage after welding. Otherwsie T6 becomes T0 and your yield strength drops to 70

#23 Rasputin

Rasputin
  • Member

  • 960 posts
  • Joined: February 10

Posted 23 September 2012 - 05:58

Very true, welding 7075 is typically not a good idea, perhaps with the xception of FSW, but the problem is also about micro-cracking.

Just like they do within the aerospace-industry, 7075-T6 should be bonded rather than welded.

I see no problem with that on a roll-cage, that's the way the entire Formula One chassis is put together, isn't it? Ideally you need an
oven for that as well, but there are also room temperature methods. I have actually made rock-bolts in 6082 that way.

#24 Tenmantaylor

Tenmantaylor
  • Member

  • 18,042 posts
  • Joined: July 01

Posted 23 September 2012 - 08:18

Steels main advantage over alu and carbon fibre is malleability, it's ability to deform before reaching yield point. e.g, it could be quite bent and damaged and still have inherent strength. Carbon fibre is capable of a higher yield strength and obviously a lot lighter with it but is a lot more brittle and has nowhere near the malleability. Alu has the lowest yield of the three but is in the middle in terms of lightness. The argument is then one of design for roll cages, is malleability desirable? Probably wouldn't be in terms of an F1 safety cell as the tight confines of the cockpit would probably result in the driver being crushed. Roll cages probably not as critical as there is more space around the driver so you'd have to assume alu makes more sense in rollcages than in formula cars at least. Can't see why a modern designed alu rollcage using modern techniques couldn't be as strong as steel. It would have to be more bulky than steel though that is for sure and a lot more expensive although a lot cheaper than CF. The car industry is a good barometer, even though F1 cars were made from alu over 40 years ago it's use is only just starting to become widespread in the automotive industry due to cost constraints.

#25 Rasputin

Rasputin
  • Member

  • 960 posts
  • Joined: February 10

Posted 23 September 2012 - 08:42

It's true that 7075-T6 aircraft-aluminium has very little elongation between yield- and rupture, xtruded pipe yields at 485 MPa (70 ksi)i and ruptures at 540 (78 ksi) with 8% elongation,
but the yield is almost in parity with 4130 steel at one third of the mass. The downside is as already said above is how to make the joints, where welding is not the recommended way.

#26 Catalina Park

Catalina Park
  • Member

  • 6,768 posts
  • Joined: July 01

Posted 23 September 2012 - 09:46

I saw an open wheel car with an aluminium hoop fall over at Eastern Creek and it slid upside down for a long time. Long enough to grind halfway through the pipe!
The best bit was it was in a class that was supposed to have an all steel chassis/roll bar. :rotfl:

#27 bigleagueslider

bigleagueslider
  • Member

  • 1,235 posts
  • Joined: March 11

Posted 25 September 2012 - 03:30

It's true that 7075-T6 aircraft-aluminium has very little elongation between yield- and rupture, xtruded pipe yields at 485 MPa (70 ksi)i and ruptures at 540 (78 ksi) with 8% elongation,
but the yield is almost in parity with 4130 steel at one third of the mass. The downside is as already said above is how to make the joints, where welding is not the recommended way.


As others noted, 7XXX and 2XXX series aluminum alloys cannot be fusion welded because they will crack. Aluminum alloys like 6061-T6 fusion weld quite nicely, and TIG welds using 4043 wire will eventually naturally age harden to a T6 condition. 4130 alloy steel fusion welds nicely, but the resulting weld HAZ will only have normalized material properties at best.

#28 Rasputin

Rasputin
  • Member

  • 960 posts
  • Joined: February 10

Posted 25 September 2012 - 06:29

Very true, TIG-welding 6000-series is no problem at all and the seam will age of course, but even in T6 condition, the yield strength will only be half of 7075-T6 at best.

I'm actually working on some custom-made M24 bolts and nuts from 7075 for an off-shore application as we speak and it's not that difficult to machine or surface-treat either.

#29 Lee Nicolle

Lee Nicolle
  • Member

  • 11,038 posts
  • Joined: July 08

Posted 29 September 2012 - 05:24

A little off subject but alloy motorcycle frames have been cracking, both around welds and in stressed areas. Generally higher km or hard worked machines but still relevant to this discussion.

#30 Rasputin

Rasputin
  • Member

  • 960 posts
  • Joined: February 10

Posted 29 September 2012 - 05:50

As discussed above, this is most probably due to welding of 7000-series Alu, a material with a yield-strength of some 500 MPa, but terrible welding-properties.

Jet fighters and airliners are typically built from 7075 or 7068, but they are bonded together today, while being riveted in the old days, rarely welded.

#31 Chubby_Deuce

Chubby_Deuce
  • Member

  • 6,870 posts
  • Joined: July 04

Posted 30 September 2012 - 03:29

A little off subject but alloy motorcycle frames have been cracking, both around welds and in stressed areas. Generally higher km or hard worked machines but still relevant to this discussion.


GSXRs specifically? If so I'm pretty sure that was a manufacturing problem, not design.