Should new convertibles have taller than roof roll bars?

[Andres125sx]

It´s not a matter of how do they look, but how do they work. A tilted pillar that must take a vertical load applied at its end will never resist as much as a vertical pillar receiving the same load.

Never??? A well designed convertible could be stonger in a roll than a badly designed saloon.

The windshield on any car will be designed to resist the applied load within the deflection criteria. If a convertible has a more slender structure then it'll have a thicker section to give it enough strength. That's what engineers do.

Yes they could, but reality is they´re not. A pillars would need to be too thick because of the explained reasons, it´s a cantilever, a more slender structure, and must resist side impacts

That´s the reason C3 Pluriel or Fiat 500 are designed with antiroll arcs, that´s the only way a convertible will be as safe as a car with a roof




Obviously they´re not as beauty as a traditional convertible, but they tried a different approach, safety first

[Andres125sx]

I dont have any statistics but when i was involved in a car crash and ended up in hospital with a bad hand injury the surgeons said convertibles cause lots of hand injuries - peoples arms can be flung up without their control and touch the road as the car flips. In my own crash (not a convertible) the airbag deploying possibly contributed to my arm smashing the drivers side window and as the car rolled it touched the road, briefly, but enough to remove a lot of skin, damage tissue and fracture bones. I guess theres not alot that can be done to stop issues like that without racecar side netting- thats in a roofed car, without a roof it wouldnt help..

I recently read the usual recomendation of holding the wheel at 10-2 (as if it would be a watch) or 9-3, should be ignored. With airbags it´s safer to hold the wheel at 8-4 because of that, if the airbag deploys it´s safer if your hands are lower

[mrluke]

Never??? A well designed convertible could be stonger in a roll than a badly designed saloon.

The windshield on any car will be designed to resist the applied load within the deflection criteria. If a convertible has a more slender structure then it'll have a thicker section to give it enough strength. That's what engineers do.

Yes they could, but reality is they´re not.

Citation needed.

Do you have any evidence to show that convertibles perform worse than tin tops?

[Greg Locock]

Indeed, I did a data mining exercise on IIHS data for crash rates between convertible versions of tin tops and their regular versions.

sadly for the chicken littles on this thread the convertibles were in general safer, that is, fewer casualties per distance driven.

So this is a thread about nothing, and it isn't as funny as Seinfeld.

As this graph makes clear, the safest thing g-force addict could do rather than worrying about other people's choice in cars is to cut his testicles off and drive like a girl

http://www.consumerreports.org/cro/maga ... /index.htm

[Cold Fussion]

It´s not a matter of how do they look, but how do they work. A tilted pillar that must take a vertical load applied at its end will never resist as much as a vertical pillar receiving the same load.

Never??? A well designed convertible could be stonger in a roll than a badly designed saloon.

The windshield on any car will be designed to resist the applied load within the deflection criteria. If a convertible has a more slender structure then it'll have a thicker section to give it enough strength. That's what engineers do.

Yes they could, but reality is they´re not.

I'm seeing a lot of hand waving but no actual analysis. I'm not sure why you don't grasp that two structures designed to meet the same standard should be expected to perform similarly. Your crusade against the cantilever is pretty misguided as well, a roll hoop is a cantilever as well, unless the car lands exactly vertically upside down (a highly improbable scenario).

[Andres125sx]

Never??? A well designed convertible could be stonger in a roll than a badly designed saloon.

The windshield on any car will be designed to resist the applied load within the deflection criteria. If a convertible has a more slender structure then it'll have a thicker section to give it enough strength. That's what engineers do.

Yes they could, but reality is they´re not.

Citation needed.

Do you have any evidence to show that convertibles perform worse than tin tops?

Yes, it is in the same reply you quoted.... the part you deleted. What would be the point of a convertible with those ugly arcs if they would be equally safe just with pop-up rolls and a windshield?

[Ciro Pabón]

A tilted pillar that must take a vertical load applied at its end will never resist as much as a vertical pillar receiving the same load.

Ehem, that's simply not true. I'm with Cold Fussion here (nice to meet you, Cold), and even not taking in account that a car that rolls is usually moving forward, so no pillars work as columns, let's assume for a moment they do.

If they are columns, then you have to take in account slenderness. Simple: usually, a slender column working in compression fails before a similarly slender cantilever does. That's why complex beams are made of inclined members with few members in a vertical position, to diminish the possibility of buckling.

A cantilever flexes more (actually, four times as much as a simply supported beam) but a slender column develops buckling because... Euler.

Euler buckling modes for vertical slender columns


Besides, you dismiss cantilevers as if they were worse than non-cantilevered beams. Again, that's not true: it depends.

For example, as opposed to simply supported beams, a properly constructed cantilevered beam (with cantilevered ends, affixed to the supports) reduces maximum moments around 45%.

If w is the load per length, L is the length of the beam, you create supports at L/6 and the simple beam is reduced to 2L/3, then you go from a maximum moment of wL^2/8 to 5wL^2/72.

Look, ma! I reduced the moment! No hands, just a cantilever!


As far as I know, there is no structural issues for rollovers in convertibles. The structural issue with them is that the chassis instead of being a box, by virtue of having no roof becomes more like a sheet of metal (the floor). This means that the chassis has poor properties for torque resistance and thus they are not as good sporting cars as more rigid structures, among other things.

Finally, you say that you are not bothered by "if it´s really needed or not". Actually, that's a crucial point in engineering. In the same way that you do not accept homeopathy as a solution instead of regular medicines (it's not ethic!) you do not (usually, of course) create unneeded solutions.

I'm not saying that rollovers are or are not an issue on convertibles, I'm only saying that it's a perfectly valid question that could be answered.

[Andres125sx]

I'm seeing a lot of hand waving but no actual analysis. I'm not sure why you don't grasp that two structures designed to meet the same standard should be expected to perform similarly. Your crusade against the cantilever is pretty misguided as well, a roll hoop is a cantilever as well, unless the car lands exactly vertically upside down (a highly improbable scenario).

A roll hoop is also a cantilever.... with a much lower span..... much better slender ratio... and a much more friendly load direction.

Take a look to ADAC rollover tests. It´s not me who said front passengers suffer more injuries, that´s the conclusions of the rollover tests. The best at this aspect is the Mini Cooper, wich is the one with the most vertical, best slender ratio and shorter A-pillars. It fails at not using pre-tensioner seat belts tough

I´m not posting youtube link because it´s explained at the description:
http://www.youtube.com/watch?v=_lE88-6TmEA

[Andres125sx]

Besides, you dismiss cantilevers as if they were worse than non-cantilevered beams. Again, that's not true: it depends.

For example, as opposed to simply supported beams, a properly constructed cantilevered beam (with cantilevered ends, affixed to the supports) reduces maximum moments around 45%.

Exactly, with cantilver ends as opposed to simply supported beams. A windshield is a simply supported beam

If w is the load per length, L is the length of the beam, you create supports at L/6 and the simple beam is reduced to 2L/3, then you go from a maximum moment of wL^2/8 to 5wL^2/72.

Look, ma! I reduced the moment! No hands, just a cantilever!
http://upload.wikimedia.org/wikipedia/c ... ge.svg.png

That´s not a true cantilever, the suspended span makes cantilever arms work different way, they start working as compressed structures while a true cantilever works as a flexed one (compresion at the lower side and traction at the upper one) with much higher moments at the support. Once you put some more support you highly reduce the moments, but that´s not the case on a convertible windshield

[Richard]

Indeed, I did a data mining exercise on IIHS data for crash rates between convertible versions of tin tops and their regular versions.

The convertibles were in general safer, that is, fewer casualties per distance driven.

That's interesting, it'd be great to see the numbers. I wonder if there is a difference between convertible versions of normal family cars compare with sportier roadsters. One could imagine the dull family convertibles will have higher mileage on the school run and family holidays so that would skew the average for all convertibles. I guess the powerful roadsters driven by reckless road hogs will have a lower mileage and more crashes? The problem wouldn't be the design of the car, its the mentality of the driver.

[Cold Fussion]

I'm seeing a lot of hand waving but no actual analysis. I'm not sure why you don't grasp that two structures designed to meet the same standard should be expected to perform similarly. Your crusade against the cantilever is pretty misguided as well, a roll hoop is a cantilever as well, unless the car lands exactly vertically upside down (a highly improbable scenario).

A roll hoop is also a cantilever.... with a much lower span..... much better slender ratio... and a much more friendly load direction.

Take a look to ADAC rollover tests. It´s not me who said front passengers suffer more injuries, that´s the conclusions of the rollover tests. The best at this aspect is the Mini Cooper, wich is the one with the most vertical, best slender ratio and shorter A-pillars. It fails at not using pre-tensioner seat belts tough

I´m not posting youtube link because it´s explained at the description:
http://www.youtube.com/watch?v=_lE88-6TmEA

What is this video from 2003 meant to show? All I see are cars which are evidently not terribly well designed for rollover safety causing bodily harm to crash test dummies. Are you actually trying to suggest that it is impossible to design a cantilevered beam to withstand extreme deflection during rollover? Because such a position is untenable.

[mrluke]

A more interesting video would show a tin top and convertible version of the same (modern) car.

[Andres125sx]

What is this video from 2003 meant to show? All I see are cars which are evidently not terribly well designed for rollover safety causing bodily harm to crash test dummies. Are you actually trying to suggest that it is impossible to design a cantilevered beam to withstand extreme deflection during rollover? Because such a position is untenable.

Where do you get that conclusion from??

We´re discussing how to improve convertible´s safety. Some people said they´re not sure if it´s even needed, and I posted that video where, as you´ve stated yourself, convertible´s safety is not that good in case of a rollover. So yes, it´s needed

Then....

But since front passengers suffer more injuries, IMO they should:

A- Reinforce windshield pillars
B- Increase lenght of pop up roll bars
C- Start using front pop up roll bars


A is cheapest and less safe, while C is safest and most expensive

And basically that´s what the thread goes about, if convertible´s safety must be improved and how to do so

[Andres125sx]

A more interesting video would show a tin top and convertible version of the same (modern) car.

Agree, now you just have to post the link

[Richard]

We´re discussing how to improve convertible´s safety. Some people said they´re not sure if it´s even needed, and I posted that video where, as you´ve stated yourself, convertible´s safety is not that good in case of a rollover. So yes, it´s needed.

What is the performance of current cars? If some of them are the same as the 2003 video then it would seem that the regulations would need to be better in terms of deflection, but we don't know that.

I deliberately use the term deflection. If the regs protect a higher headspace then a manufacturer can increase the height of the windshield to provide an extra crumple zone. The shield can still absorb energy and flex, it would protect someone as tall as a dummy (188cm). It would also help sell cars to taller people due to the increased headroom. Those taller people wouldn't benefit from the tougher regs because their heads are in the crumple zone.

I suppose the ultimate safety would start with a taller dummy say 210 cm and fill the crumple zone with a soft material so the headlining is on the edge of crumple zone. That would stop occupants who sit too high for the designed deflection.

Of course this is all abstract musings, no one has identified the current test requirements, or how many people suffer head or vertebrae injuries in rolls.