Ever since the IIHS started doing crash tests, some vehicle makers have been cheating. What I mean by this is they do things to make the vehicle perform well on the test, but not necessarily to do much better, or in some cases to perform worse, in more diverse real world accidents.
Here is a picture of what they did on the 2019 Nissan Altima. Since the late 1990s most cars have had short width bumpers or collision strips as they are often called in the parts list. They say it is to save weight, but I suspect allowing the corner with the headlight unprotected and easily damaged makes a lot of money in the spare parts market. I believe having the bumper chopped off with no structure behind it doesn’t help with the newer small overlap crash tests. Some 1997 to 2012 vehicles may actually do worse in small overlap crashes than older models.
In this case they extended the sub frame, or engine cradle, out to the sides. This would help the IIHS small overlap test, and it would help if hitting a light pole. But the top of this thing is only 11.5" above the ground, too low to contact the bumper on another car. In a real world crash with another vehicle it doesn’t help. On the 92-96 Camry there is a narrow bumper, but it has a structure kind of like this at the edges of the bumper to widen it. The Volvo 850 and 2001+ V70 platform make the unibody frame wide at the ends. Two examples of this being done correctly.
And here you go again, worrying about the cost of repairing the vehicle and not the life that was saved.
1 Like
It doesn’t even do that, if the crash is with another vehicle.
The bumper bar is very light weight, the short width is part of the collision design. A full width bar might result in a harsher impact, cause the vehicle to rotate and roll-over. There is also consideration for pedestrian impact protection.
2 Likes
It is normal part of design process to go through a hazard and risk assessment that lists every conceivable hazard for the product and calculates a risk factor based on severity and probability of occurrence. Then what mitigations are developed to lower the risk factor. There are always trade-offs. Risk of human injury always trump material considerations, of course.
But the thread was started with the implication that products may be designed to pass the test. That is certainly true. And a solid barrier 5 feet tall does not act the same as impacting another car. But they have to meet the test criteria and one might argue they should not second guess the motivation behind the engineered test. That could lead to dangerous assumptions.
Many years ago, I worked on a project with an optical window and what the regulatory bodies define as a “high voltage” inside the enclosure. Optical windows are often fragile and break easily upon impact. Much engineering can go into hardening these optical “glass” windows, similar to what you find in corning glass products like a coffee pot that can withstand extreme thermal swings as well as impact resistance.
But the safety test is very specific- a sphere of a given weight and dimension is swung from pivot point at a specific angle and length into the window.- It is almost impossible to pass this test with optical windows like borosilicate or fused silica substrates. So what do you do? You design an aperture slightly smaller in size and recessed enough that the sphere can never reach the window. Now it passes the test. Does it protect against all hazards in the field? Probably not. But then again, I am not positive what went into the thinking on the test design either… 
1 Like
Then the upper management decides to overrule all of this for “cost saving” or some such excuse!
Yes the ball test is a good example of designing for a test. When the IIHS moderate overlap test first came out, the cars at the time that did well had good overall safety, since they were not yet tweaked to do well on this specific test. Note that good = will cost insurance less money, and poor = will cost insurance the most money.
