Crush distance and crash survival
There are a lot of ways to have a dangerous impact in a race car.
1. One is to have a small object flying towards the driver.
...a. If it is very small, a helmet or canopy might stop it.
...b. If it is big, a roll cage might stop it.
2. Another is to collide with something large and heavy enough to slow the car down.
...a. A car with a very rigid frame might keep the object away from the driver.
.......But it will cause very high deceleration rate which may harm the driver.
...b. A car with a crushable frame might reduce the g forces on the driver.
.......But it may allow the intruding object to get to the driver.
At high speeds, you need a lot of distance to slow down or stop at safe levels of deceleration. So a crushable frame is preferred. The frame has to be designed to decelerate the car fairly smoothly. It doesn't do any good to have the first two feet crush with no resistance at all, and then have the last inch take the full brunt of the impact. If the driver decelerates at more than 100 g's, the chances of serious injury are high.
Assume that the vehicle provides fairly constant deceleration in a crash. The crush distance required increases as the square of the impact velocity. If you don't have enough crush distance for a given speed, your deceleration is going to be very high, regardless of the materials and design of the car. All the car can do is absorb energy in a smooth manner so as to give smooth deceleration.
Not showing all the math here.
Remember, this is just to limit the deceleration to 100 g's, which is going to be very painful. If you crash at 20MPH, the safety harness might give you enough crush distance (1.6 inches) to keep you safe, even if the car is completely rigid. At 80MPH, you will need two feet of crush distance. Hopefully, the car has that much, and absorbs energy smoothly over that distance. At 120MPH, you need five feet of crush. You might get that if you hit a tire wall or another car with a crushable frame. But it's getting iffy. Above 120MPH, you really don't want to decelerate to a stop in a crash. For example, at 200MPH, you need 13 feet of smooth deceleration. That is not likely.
What if a rigid part of the car, like the roll bar, strikes a stationary object? It will try to decelerate instantaneously to zero speed (not taking into account that it is slightly elastic). But it is attached to a 1600 pound car that doesn't want to stop instantaneously. The forces could be enough to tear the roll bar out of the car. So placing a roll cage around the driver will help if foreign objects are flying, but it won't help absorb energy in a big collision with a wall or car. Let's say you could build a roll bar that could withstand unlimited forces. The driver might decelerate at 1,000 g's and end up dead anyway. He will be crushed by the safety harness or the seat, or his brain will collide with his skull. You always want to provide some crush distance, no matter what angle the car collides. The problem with very high speeds is that you can't possibly provide enough crush distance.
If your body is being slowed down by the safety harness, you want your head to be slowing down by the same rate, or you could end up with spinal cord injuries. This is what the Hans device does. It stiffens the connection between your head and your torso. http://en.wikipedia.org/wiki/HANS_device But even the Hans device cannot withstand unlimited g forces. Some vendors make outrageous claims, like "Reduces Impact forces by 80%". No, all it does is keep the head aligned with the torso. The only thing that can reduce impact force is more smooth deceleration distance.
If you want to understand helmets, canopies, roll cages, Safer barriers, tire walls, etc., you need to understand crush distance.
Last edited by Motie; 09-03-2012 at 07:51 PM.
so is a side impact actually worse than head on impact if both stop the car?
what does safer do to your math?
Safer provides crush length, but it's not enough for a head-on crash at 200MPH. It will cushion a glancing impact. I'm not an expert on Safer, but it looks to me like it absorbs energy; it doesn't act like a spring, where the energy goes right back into the race car and hurls it back out onto the track. Think of the car's velocity as having two components or vectors: one in the direction of the track, and one at a right angle to the track. What if you are moving at 180MPH in the direction of the track, and heading towards the wall at 60MPH because of a collision with another car? According to my graph, at 60MPH you need about 15 inches of crush distance to limit the deceleration to 100 g's. Between the Safer and the car, you should have at least 15 inches, so you will probably survive. If the wall was concrete, you would only have the crush length built into the car. The Safer is not going to absorb all the energy due to your 180MPH track speed, but it will slow you down somewhat as it deforms.
Originally Posted by ndcrs
Side impact: it depends on how the car is designed. The sidepods will help. But I don't know how the human body responds to side impact. Does the Hans device work on side impacts? I don't know.
It's important to remember that crush length is not enough. The car or the wall needs to crush in a way that absorbs energy (a LOT of energy) at a fairly constant rate. If you've got 2 feet of crush length, but all the deceleration comes in the last inch, then really you've only got one inch of crush length. Fiberglas bodywork looks big and safe, but it doesn't absorb energy; it just collapses into a million pieces.
What if something heavy and inelastic hits the side of the cockpit just below the driver's head, like in the Bourdais crash? That's really bad, because there is a lot of energy and almost no crush distance before the driver is struck. What if you put an indestructible roll cage out there? A huge amount of momentum is going to get transferred from the heavy object to the roll cage, which means that the car is going to bounce off the impacting object at a very high rate of acceleration. If it's higher than 100 g's, the driver may be severely injured. Or the roll cage might detach itself from the car. The only solution is to build in crush length. An enclosed LMP1 car will have some crush length all around the driver's head, because the roof is extremely strong. But if the impact occurs on the roll bar, it's not crushable, so the g forces on the car and driver will be very high.
It's an extremely challenging problem, even if you had unlimited money to throw at it.
This well presented thread illustrates the issues with just grafting a 'top fuel style' rollcage into an IndyCar as many have suggested. I think too many people simply look at steel tubing and instinctively think it has to be stronger and better than a carbon fiber tub.
Thanks for posting this.
I don't mean to hijack this thread or anything but I got to thinking after looking at it about the angle of impact. At first glance I thought 60mph perpendicular to the wall seemed a bit high. I got the calculator out and using your numbers I got an angle of impact of 18.4 degrees. So that got me to thinking about what the angle of impact of a car at Indy might be if a car were to go straight-on into the wall.
Using Google Earth (I've got the pics if you are dying to see them) and Photoshop I was able to come up with a rough guess-timate of 17.2 degrees for a car going straight on from mid-corner into the wall. Using Google Earth's historical imagery I was also able to come up with a guess-timated angle of impact of 21.1 degrees for a car running down on the old aprons before they put in the warm-up lanes.
With these two angles we can now see just how much of a difference those 4 degrees or so can make in a crash situation. For a car going 225 on the old configuration it would hit the wall at 81.0mph and on the new config at 66.5mph. Fourteen and a half mph might not seem like a huge difference until we calculate the difference in kinetic energy that is going into the wall. For those who forgot, kinetic energy is .5*mass*velocity*velocity. Using metric units and a car weighing 765kg (current car weight) I came up with a KE=501,523N-m for the old configuration and for the current config, a KE=338,309N-m. (For those who are curious, for a head-on crash the KE=3,869,806N-m -not counting the rotational energy of the tires) So this tells us that by altering the angle of impact those by those 3.9 degrees we have reduced our impact energy by 37%. Of course real life is never quite so simple but it does give you at least a rough idea.
What if two cars come together at racing speed? There isn't much difference in speed between them, so there won't be big g forces on the driver. But the driver can be injured if his head gets hit by the other car. The roll cage will keep the big parts of the other car away from the driver. Or what if the car goes into the tire wall? The roll cage will keep the tires away from the driver's head. The tire wall provides crush distance, so the g forces won't be too bad, even though the roll cage is rigid.
Originally Posted by rrrr
You can make pro and con arguments about canopies and roll cages, but the thing to remember is that no one thing is going to protect the driver in all possible scenarios. Also, if you crash into a rigid surface roof-first at high speed, I can't think of anything practical that's going to save you. How are you going to add crush distance to the roof? Sure, you could enclose the cockpit LMP1 style, and add a lot of crushable metal on top, but it would look ridiculous.
Last edited by Motie; 09-15-2012 at 12:36 PM.
I assume that somewhere there is a prioritized list if significant accidents, but I've never seen one. What that would show is what you need to protect against. It's impossible to effectively counter every possibility, but I think people tend to respond to what brings the headlines. Dan Wheldon's accident is quite rare, I think. The only other one similar that I remember was Greg Moore's accident. Anyway, I'm not suggesting to not work on overhead protection, but should that be the first priority?
conway (both ends of indy), briscoe, rice, meira dario were all basically the same with accident as dan's without the car being rotated a small amount. there are more that i could come up with in 5sec. i think the op is suggesting impacts more like bourdais at sonoma or a stan fox type accident.
Originally Posted by flatlander_48
I was speaking of accidents where the car is on its side or completely inverted and the top side of the car strikes something. I checked the Conway, Briscoe and Rice accidents. Conway and Briscoe hit the fence bottom first. Rice was inverted, but completely on the track. He never hit the fence. I didn't take the time to look at the other 2.
Originally Posted by ndcrs
My point was that, if you want to make improvements, you have to understand what the real hierarchy of problems is. Usually, you can't work on everything at once, so you have to decide what to tackle first, second, etc. I would assume that someone has done a statistical analysis of the accidents in the ICS, but I don't think it has ever been published. Crush distance and impact mitigation may not always be the major factors in accidents. For example, a flip and roll where you don't hit the wall or fence may or may not involve the roll bar hitting the track, but energy dissipation as suspension pieces are shed can also be important. Anyway, it's one thing to solve problems, but it's equally important to make sure that you are working on the right problem.