IIHS crash testing on small cars

Messages
2,147
Location
Chicago, IL
Guess that side curtain airbag really makes a difference. I think I saw somewhere on bitog recently where a pothole cracked a wheel in two in the same way. Someone mentioned that it's more common now with the high offset wheels, there ends up being a big lever arm since the inner lip is so long.
 
Messages
11,660
Location
NorthEast
Being an engineer, I can only say this:- There is no way in the hade any of these tests are going to be reproducible to any lengths whatsoever. Imagine a single degree difference in the approach of the vehicle and the results would be completely different. If the car approaches the wall at 89.5 degree instead of 90 degree, you would get completely different collision dynamics. I do not believe the contraption which conducts this crashes has the ability to keep the vehicle approach to the wall perfectly identical in each of its run. - Vikas
 
Messages
1,797
Location
Texas
It looks like the difference between airbag or not in this instance is open casket or closed. How many vehicles are there that can sufficiently protect the driver in this kind of crash? My understanding was that an offset wreck such as this at any normal speed basically means death or serious injury.
 
Messages
16,043
Location
N.H, U.S.A.
I think the cx-9 did worse. Are not these old Ford joint venture platforms ( not the new independent Mazda Syyactive platforms? Ill look it up .... The MItsubishi Oullander did VERY, VERY well in the quite severe small overlap test.
 
Last edited:
Messages
40,736
Location
Great Lakes
Originally Posted By: Vikas
Being an engineer, I can only say this:- There is no way in the hade any of these tests are going to be reproducible to any lengths whatsoever. Imagine a single degree difference in the approach of the vehicle and the results would be completely different. If the car approaches the wall at 89.5 degree instead of 90 degree, you would get completely different collision dynamics. I do not believe the contraption which conducts this crashes has the ability to keep the vehicle approach to the wall perfectly identical in each of its run. - Vikas
Looks like they have various markings on the vehicle and on the stationary object and then a camera that's probably capable of detecting if it was hit at the right angle. Just wondering how many cars they had to destroy before the got the angle right. Still, what's the likelihood that when you crash, you'll do it at exactly 90 degrees and at exactly this amount of overlap? Another words, how representative are these tests (and test results) of real life situations? And we did have this discussion regarding this small overlap test before, so I'll leave it at that. In the end, you can build a car to ace any test. The question is: what is the cost to the end customer and what is the benefit? And who benefits if a car does not get a perfect score?
 
Messages
119
Location
texas
As for this crash test not being valid, I disagree. Sure you can look at the NHTSA crash tests ( safercar.gov ) to combine with iihs.org crash tests, for the big picture. Bottom line, stronger A-pillars and floor pan elements, crush zones near the engine, etc. all contribute.
 

Hokiefyd

Thread starter
Messages
14,505
Location
Top of Virginia
Any given test will truly represent only a fraction of real-world variables that happen in crashes. Few crashes probably occur the way any crash test is conducted. They're done under controlled environments and we know the world is very much uncontrolled. But then what is the alternative? Never crash anything together to see how body structures and restraint systems could be improved?
 
Messages
9,521
Location
Canuck - moved to Texas
Those tests obviously have some merit in real world safety improvements; otherwise we would not have strategic crumple zones, air bags and self-tightening belts just to name a few examples. Obviously we cannot test for every possible scenerio, but don't forget that these tests also supply very important data for the simulation models used in CAD.
 
Messages
11,660
Location
NorthEast
My point is that there must be quite a lot of variability in the end results even in the controlled crashes i.e. a very small change in the control parameter is going to cause a huge change in the crash damage. But I do agree that it needs to be done and that is the only objective way to do comparison even with its shortcomings.
 
Top