Closed Car Roll Over Research – Full Story

There’s Going To Be An Accident…

Months of preparation go into the precisely calibrated collision that produces the twisted, broken chassis you’re about to see. This Subaru will slam into this test rig at the Millbrook development facility under the watchful eye of FIA Institute research consultant Andy Mellor, who is here testing potential improvements to the roll-over protection system fitted to competition rally cars.


A blustery November day in the middle of England. But, as autumn stares winter in the face in this corner of Bedfordshire, work goes on. The scene looks similar to a James Bond film set, with cameras, laptops and levels of technology that would impress Q. There’s final fussing over four ordinary-looking Subarus covered in sensors, accelerometers and potentiometers. One of the cars is now placed on its side on a flatbed trolley. Facing a wall.

An accident is waiting to happen. More nervous glances. More talking into handheld radios. In three, two, one…

The car is fired, roof first, towards its impact target. Only the whirr of the rails, as the flying floor runs along its tracks, breaks the silence. Then, bang. The silence is well and truly shattered.

Mellor inspects the set upWelcome to the FIA Institute’s latest round of ground-breaking research into Roll Over Protection Systems (ROPS). Having long poured over post-crash data, the FIA Institute is immersing itself in a more hands-on approach, making detailed studies of the component parts of roll-cages and what constitutes the optimum protection for competition cars. And now those investigations are taking the researchers to the very heart of the accident.

A roll-cage is relatively self-explanatory: it protects a car’s occupants in a roll-over accident. To do that, it has to be exceptionally strong. But there’s more. And today, at the Millbrook proving ground, AUTO is learning just how much more there is.

FIA Institute research consultant Andy Mellor frames the question that explains the tests: “Should a roll-cage be strong or energy absorbing? In fact, it needs to be both – a compromise. But metallics tend not to behave like that – they’re either strong or stretchy.”

All four cars are fitted with roll-cages, each built from a different type of steel: first, T45 – a chrome-manganese, aerospace-standard metal; then 15CDV6 – a chromium-molybdenum-vanadium product reckoned to be the strongest and potentially the most suitable for welding; ROPT 510, which attempts to combine strength and ductility; and finally CDS, the most economical and ductile alternative.



The cars have been prepared by Prodrive, one of Britain’s most successful and respected motor sport firms. Prodrive technical director David Lapworth is watching his work hit the wall.

The first impact, centred mainly on the Subaru Impreza’s windscreen A-pillar, causes significant bending and damage to the T45 roll cage. Mellor, Lapworth and the rest of the team descend on the car immediately after the crash. Notes are taken, data logged, stats stored for later deep analysis.

“That first accident was a reconstruction of a serious, but not catastrophic crash,” says Lapworth. “That was a roll, if you like, with the car impacting the ground at 30km/h and the top of the A-pillar taking the full force. We’re looking to replicate the vertical drop of the car. This isn’t about a big, dramatic, head-on impact; it’s about the roll-over protection system. When a car rolls, it rarely drops

off a cliff and lands square onto its roof. Think about firing a gun horizontally and dropping a bullet from the same height, the bullets would hit the floor at the same time, and with the same vertical speed. That’s what we’ve got here.”

Mellor adds: “This is about the resistance of the roll cage to both force and energy. It’s not a static load test.”

In the second crash the 15CDV6 roll-cage is put to the test. From the outside the deformation looks similar. A quick glimpse inside the car reveals significantly different deformation patterns. And the Institute’s use of cutting-edge technology to monitor stress, stretch and shift levels in the cage is at work. The physics of what just happened is being calculated on a raft of laptops, all of which are readying reams of data aimed at reducing the danger involved in future roll-over accidents.

Mellor is looking over shoulders, staring at screens. Both he and Lapworth have been working on this project for two years and
are impatient to see the results. They want to see the science. But they must wait. We’ve just seen the performance of the two high- performance steels (T45 and CDV6) typically used for World Rally Cars. Next will be lower strength but more ductile materials often targeted at club racing (ROPT 510 and CDS).

Mellor leans in and inspects the damage from the crashes.

“We’re looking for strong materials with compliant joints which will deform,” he says. “Ideally, we want the cage to deform by up to 400mm without any joints in the cage failing.”

The cage itself is only part of the equation. What binds the sections of steel tube together is the weld – which raises the perennial question of the best welding solution. Is it TIG (Tungsten Inert Gas) or MIG (Metal Inert Gas)? Today will also go some way to providing that answer.


A series of near identical and precisely angled collisions in simulated roll-over
accidents test the resilience and strength not only of the steel used in tubular
cages fitted to competition cars, but also the quality and nature of the welds.
The impact target is on the right. 


“When a car rolls, it rarely drops off a

cliff and lands square on its roof”

David Lapworth



The optimum solution – if such a thing exists – is a cage that, in Mellor’s words, is both strong and stretchy.

“During a side impact into a tree, you want ultimate energy absorption,” says Lapworth. “Controlling instrusion is desirable, but energy management between the driver and the tree is critical. But during a roll-over accident you must limit intrusion of the roof to avoid direct contact with the driver’s helmets. This requires both strength and energy absorption from the cage. Some degree of deformation of the cage and intrusion towards the drivers is unavoidable in order to manage energy, but the distance between the point of impact and the crew inside must not be exceeded.”

Another major benefit of being under the skin of an accident in this way is the potential for further developing the roll-cage design.

“Once we have all the results,” says Mellor, “we will go away and look at the [ROPS] design and most likely take that on to the next phase of research. We have to remember the conflicting requirements of ROPS: it has to protect those inside, but at the same time provide space for emergency extraction of the crew; and also allow for the possibility of rescue teams needing to cut through the tubes with the jaws of life after an accident.

“Today we’ve had a unique look inside an accident, which makes this pioneering work. What we’ll study now is the sequence of failures within the roll-cage. For example, when the triangulation bar fails, the cage can lose its structural stiffness. So replacing those joints with something more elastic or tolerant may allow more bending in the joints while keeping the whole structure more intact.”

But that’ll be for another day and another accident.

Having spent much of his working life overseeing roll-cages going into race and rally cars, Lapworth is as experienced in this field as they get. He is keen for the results from this test to further help define the FIA’s current technical regulations regarding ROPS.

“The regulations have evolved over time,” he says, “which is only natural with so many new formulas and categories coming in, but what we now have the chance to do is look at rewriting that regulation definitively.”

Now, though, in the gathering gloom, there’s a sense of a worthy and worthwhile job well done. Never has so much hot metal been scrapped in such a good cause.

Rops pic 3

Rops pic 6



For Further Information:

FIA Institute link