Logic Problem: Assessing the Damage
09-14-11, 11:16 PM
#1
Out of Warranty
Thread Starter
Logic Problem: Assessing the Damage
Back in WWII bombers were coming back from missions over Germany, often badly shot up. The RAF asked Abraham Wald, a British statistician, to investigate the damaged aircraft and assess where new armor should be placed to protect it from flak damage. He did so rather quickly and the result was surprising. He directed that armor be placed in those areas that HADN'T been hit.
The logic goes like this. Obviously the aircraft that were returning to base had survived, while those that did not were assumed to be destroyed. In those days anti-aircraft fire was rather inaccurate as targeting was very poor, resulting in a random distribution of flak and projectiles. If the returned aircraft survived the damage they had, those that were shot down could be assumed to have been hit elsewhere, in places more vital - making a good case for armoring everything that hadn't been damaged because it was evident the bombers could still get home when shot up in less critical areas.
What can we learn from this, and can we apply this methodology when assessing random damage to cars, homes, and even ourselves? We know for example, that in most relatively minor auto accidents, the front end of the car takes the majority of the damage. What causes an automobile to be undrivable following a fender bender? We can name a few common characteristics, damage to the radiator, steering and front suspension, and the deployment of airbags in the interior. Seldom is a car rendered unusable by damage to the engine or transmission, the braking system, or electrical system. The expensive bits survive pretty well, it's the relatively cheap stuff up front that gets whacked. It's just that there's so much of it.
Why then do we not concentrate on the damage that may be less catastrophic? Rather than engineer crush zones that sacrifice the cooling system, steering gear, and premature firing of the SRS in relatively minor collisions, why not expend a little engineering time on locating these critical components away from the places that can be expected to take the brunt of a blow and collapse in a relatively minor accident. How many times have you seen a perfectly good automobile totaled when the vast majority of the collision damage is forward of the radiator bulkhead - and the rest of the car is completely undamaged. Sheet metal is cheap, but once the force of a collision reaches key structural members, the day is over. The car will literally tear itself to pieces to protect the occupants. That's fine at 40+ mph, but so many accidents take place at a fast walking pace.
Once upon a time when 5mph bumpers were required, we saw relatively little damage from a crash up to that speed . . . but that same crash at 6 or 8mph could be devastating. Once the engineered safety systems are stroked to their max, you start to destroy key structure. As an alternative, why don't we protect the mission-critical stuff and sacrifice the cosmetics? Here's where those plastic body panels that either bend or snap off for replacement might be helpful.
The logic goes like this. Obviously the aircraft that were returning to base had survived, while those that did not were assumed to be destroyed. In those days anti-aircraft fire was rather inaccurate as targeting was very poor, resulting in a random distribution of flak and projectiles. If the returned aircraft survived the damage they had, those that were shot down could be assumed to have been hit elsewhere, in places more vital - making a good case for armoring everything that hadn't been damaged because it was evident the bombers could still get home when shot up in less critical areas.
What can we learn from this, and can we apply this methodology when assessing random damage to cars, homes, and even ourselves? We know for example, that in most relatively minor auto accidents, the front end of the car takes the majority of the damage. What causes an automobile to be undrivable following a fender bender? We can name a few common characteristics, damage to the radiator, steering and front suspension, and the deployment of airbags in the interior. Seldom is a car rendered unusable by damage to the engine or transmission, the braking system, or electrical system. The expensive bits survive pretty well, it's the relatively cheap stuff up front that gets whacked. It's just that there's so much of it.
Why then do we not concentrate on the damage that may be less catastrophic? Rather than engineer crush zones that sacrifice the cooling system, steering gear, and premature firing of the SRS in relatively minor collisions, why not expend a little engineering time on locating these critical components away from the places that can be expected to take the brunt of a blow and collapse in a relatively minor accident. How many times have you seen a perfectly good automobile totaled when the vast majority of the collision damage is forward of the radiator bulkhead - and the rest of the car is completely undamaged. Sheet metal is cheap, but once the force of a collision reaches key structural members, the day is over. The car will literally tear itself to pieces to protect the occupants. That's fine at 40+ mph, but so many accidents take place at a fast walking pace.
Once upon a time when 5mph bumpers were required, we saw relatively little damage from a crash up to that speed . . . but that same crash at 6 or 8mph could be devastating. Once the engineered safety systems are stroked to their max, you start to destroy key structure. As an alternative, why don't we protect the mission-critical stuff and sacrifice the cosmetics? Here's where those plastic body panels that either bend or snap off for replacement might be helpful.
09-15-11, 01:18 PM
#2
I think it's a matter of economics.
Engineering new methods may cost more to manufacture than tried-and-true methods (like how expensive the compact-design engineering has been for SMART or the IQ). I see such cost cutting outcomes to be passed from the manufacturer to the consumer, as repair costs and insurance claims.
And in the realm of safety, I doubt that manufacturers will risk the potential of making the SRS too lax, even with minor accidents. A totaled/salvaged car is worth less than litigious people with expensive life-long medical bills.
Engineering new methods may cost more to manufacture than tried-and-true methods (like how expensive the compact-design engineering has been for SMART or the IQ). I see such cost cutting outcomes to be passed from the manufacturer to the consumer, as repair costs and insurance claims.
And in the realm of safety, I doubt that manufacturers will risk the potential of making the SRS too lax, even with minor accidents. A totaled/salvaged car is worth less than litigious people with expensive life-long medical bills.
09-17-11, 10:45 AM
#3
Lexus Fanatic
Originally Posted by Lil4X
What can we learn from this, and can we apply this methodology when assessing random damage to cars, homes, and even ourselves? We know for example, that in most relatively minor auto accidents, the front end of the car takes the majority of the damage. What causes an automobile to be undrivable following a fender bender? We can name a few common characteristics, damage to the radiator, steering and front suspension, and the deployment of airbags in the interior. Seldom is a car rendered unusable by damage to the engine or transmission, the braking system, or electrical system. The expensive bits survive pretty well, it's the relatively cheap stuff up front that gets whacked. It's just that there's so much of it.
Why then do we not concentrate on the damage that may be less catastrophic? Rather than engineer crush zones that sacrifice the cooling system, steering gear, and premature firing of the SRS in relatively minor collisions, why not expend a little engineering time on locating these critical components away from the places that can be expected to take the brunt of a blow and collapse in a relatively minor accident. How many times have you seen a perfectly good automobile totaled when the vast majority of the collision damage is forward of the radiator bulkhead - and the rest of the car is completely undamaged. Sheet metal is cheap, but once the force of a collision reaches key structural members, the day is over. The car will literally tear itself to pieces to protect the occupants. That's fine at 40+ mph, but so many accidents take place at a fast walking pace.
Why then do we not concentrate on the damage that may be less catastrophic? Rather than engineer crush zones that sacrifice the cooling system, steering gear, and premature firing of the SRS in relatively minor collisions, why not expend a little engineering time on locating these critical components away from the places that can be expected to take the brunt of a blow and collapse in a relatively minor accident. How many times have you seen a perfectly good automobile totaled when the vast majority of the collision damage is forward of the radiator bulkhead - and the rest of the car is completely undamaged. Sheet metal is cheap, but once the force of a collision reaches key structural members, the day is over. The car will literally tear itself to pieces to protect the occupants. That's fine at 40+ mph, but so many accidents take place at a fast walking pace.
Once upon a time when 5mph bumpers were required, we saw relatively little damage from a crash up to that speed . . . but that same crash at 6 or 8mph could be devastating. Once the engineered safety systems are stroked to their max, you start to destroy key structure. As an alternative, why don't we protect the mission-critical stuff and sacrifice the cosmetics? Here's where those plastic body panels that either bend or snap off for replacement might be helpful.
As for the cooling systems, the problem is that, in order to get cooling-air into the system (even with today's liquid-cooled engines, as air-cooled engines are pretty-much gone) the grille/radiator assemblies have to be located near the edge of the car, where air-flow is easily accessable. To cut off the cooling system from impact-damage often means simply not getting enough air.
That, for example, is one of the reasons why mid-engined cars often tend to run hot. The radiator, grille, hoses, etc... (and the cooling-slots in the bodywork) are behind the driver, but still located away from both the front and rear ends of the car. That limits the amount of air coming in, even through the slots.
Thread
Thread Starter
Forum
Replies
Last Post
CLRH2O
SC- 1st Gen (1992-2000)
11
07-02-07 11:26 PM
