Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!





Subaru Legacy/Outback 2005+

1155156158160161534

Comments

  • It's just astounding that two tests could yield such different results: it's the safest care ever in Australia!!!; it's a death trap in the U.S. Go figure. I wonder if the differences are the result of random "noise," or if instead one testing method is superior to the other. I wish Subaru would enlighten us on this.
  • sdufordsduford Posts: 577
    You shouldn't equate these results to being "the safest" nor a "death trap". Those tests only give you a basic idea of the safety of particular model under a very specific scenario. They by no mean relate to the real world.

    In fact, most car companies design their cars to do well in the tests, not necessarily to do well in the vast set of cirumstances encountered in the real world. The notable exceptions to that rule are Volvo, Saab, and M-B, all three companies spend a lot of money designing and testing for a wide set of circumstances that go well beyond the regular tests.
  • njswamplandsnjswamplands Posts: 1,760
    the car sucked at this test. period. it seems every car manufacturer who sucks at this test says it meets all standards and is safe and the iihs test is a narrow test of the possibilities for crashes.

    fact is, until an independent group crashes them in all those possibilities, this is all we have to go on.

    SO THEY DO COUNT
  • Interesting posts on safety. I thought I would compare the test methodologies (too much time on my hands) and this is what I found

    If I am correct - there are 3 identifiable differences between the ANCAP test (Australia) and the IIHS test (USA) for side impacts:

    1) 950kg trolley (Aus) v 1500kg trolley (USA)
    2) The Australian test combines the scores into 1 rating, whereas the US test provides separate front and side ratings (although you can see the split scores if you follow the links below - and the Sub scores maximum points for side rating in the Aus test anyway)
    3) Infant and toddler dummies in the back (Aus) v 12 yr old child size (USA) - irrelevant because it seems there are no issues re back seat passengers

    Some other possible differences:

    4) It is possible that a different sized driver dummy was used as the IIHS test specifically talks about smaller driver dummy - the worst results were torso and pelvis - maybe for a taller individual that would be stomach instead = reduced impact injuries
    5) It is possible that the test barriers hit the car at different heights. The impact point could also be further along the car, but the pictures suggest the same location
    6) Could there be any difference due to ANCAP cars were Japanese assembled, whereas the IIHS cars were presumably US assembled? SOA obviously thought the airbag component provider was the issue, therefore the recall. Maybe investigations could prove the steel/aluminium is manufactured to different tolerances? Before anyone gets too excited - that is unlikely.

    ANCAP methodology is listed here:
    http://www.aaa.asn.au/NCAP/explanation.htm

    Full Aus test details listing compressions scores (for comparison with IIHS test)
    http://www.aaa.asn.au/NCAP/ozindex.htm#L-MCars

    The actual numbers don't compare very easily, suggesting different methodologies or the noted differences above (especially the weight?) are the reasons.

    I don't think we'll ever get to the bottom of it - and I suspect SOA is as confused as the rest of us as to why the scores are so different.

    - Aussie Outback
  • c_hunterc_hunter Posts: 4,487
    I will try to spend some time reading over the various test procedures and maybe give some comments from an engineering standpoint.

    One major flaw in all the tests is that they do not take enough data points to have a statistically valid sample. It may be that the variation amongst many cars of the same type is low, but I have not seen any indication that they test enough cars to even come to that conclusion. I'd be very curious to know how much the results change from car to car if, say, 25 Legacys were tested.

    It would really suck if we made comparisons between different models of cars with only 1-2 data points per car. Until someone can prove to me that we do get a fair representation of a vehicle based on only 1-2 tests, I really question all of the results. If the IIHS is serious about the work they do, they must have addressed this at some point.

    Also, given the large number (near infinite I imagine) of possible accident scenarios, I would really like to see how cars handle crashes with various ranges of angle, height, velocity, mass, etc... What if the Legacy was superior from a different angle or height and the other cars were poor? It's hard to make valid conclusions based on very limited data when trying to predict the performance of any machine under a wide range of conditions. What if somebody wrote off the Legacy because of the IIHS tests, and then bought another car that had other (unknown) weaknesses that were manifested in an accident??

    Craig
  • bblachabblacha Posts: 160
    The car sucked at this test, sure. So do most cars currently. It's like 10 years ago when offset crashes were getting attention and "Marginal" would have been a distinction.

    Having said that, the Forester side of the fence looks safer all of a sudden...
  • alpha01alpha01 Posts: 4,747
    I would agree with those that have observed that the IIHS side impact test result is getting blown a bit out of proportion, but on the same token, Subaru should be paying attention to the reactions of its prospective customers.

    It is important to kee in mind that the test itself is incredibly demanding, representing a 3000+ lb. SUV crashing into the side of your vehicle. All things being equal, it would certainly be best to be in a vehicle that performs well. For me, the bottom line in this test shows that Subaru needs to head back to the drawing board with respect to the design/deployment of the side airbag mounted in the seat. The head curtain is doing its job, judging from the injury measures, and the structure held up well.

    Personally, I dont know why anyone even considers the NHTSA side impact test as a basis for safety evaluation. The star rating doesnt even include the head injury criterion, for Pete's sake (whoever Pete is...). Thus, a car scoring 4 or 5 stars in the side impact may actually inflict head injury on the test dummy in excess of 1000 (the threshold for severe or possibly fatal injury), without its star rating being affected.

    c_hunter- I understand your comments regarding testing multiple idential vehicles, but really, the concept in question in relation to structure in particular is the repeatability of deformation patterns, as opposed to the sheer varience in terms of mm, for example, that the B pillar was pushed inward, as long as that varience is within reason. I suppose I take a contrary view- unless someone proves otherwise, I cannot think of a good reason why one mass produced sample would not be significantly representative of another mass produced sample, within an accepted range.

    I remember watching many episodes of Dateline NBC in which repeat crash tests were shown, for example, and in all of them- the last gen. Infiniti Q and the Dodge Neon come to mind... the repeatability in deformation patterns of the vehicle crash structure was remarkable.

    ~alpha
  • gentlegentle Posts: 7
    I got one in the mail today.

    I got one from Ken already, so mine is up for grabs. I'm really not sure if people are still interested, what with the crash test results hogging up forum band width :)

    But if someone is still planning on purchasing this death trap, first one to email me their address gets the coupon and one foot in the grave.

    drew
    rsatica@<y a h o o>. com
  • gentlegentle Posts: 7
    Coupon is gone.

    drew
  • tsytsy Posts: 1,551
    I think death trap is a little melodramatic. It's true the results are disappointing, but I'm trying to analyze it rationally rather than emotionally. I've been looking at the technical data (what little there is) and it seems the structure held up well and intrusion into the passenger compartment was the least of any car tested by far. This tells me the safety cage is indeed quite strong and did what it's supposed to. The head was well protected, which is the real purpose of this test (read the fine print under description of the side impact test) This is why they use a 5th percentile female crash dummy, because small women are more likely to be hit in the head by the front of the SUV (if you look at some of the pictures of other cars tested you can see an indentation in the deformable barrier were the dummy's head hit- ouch!).

    Now, not that I'd want trauma to the chest, but it sure beats instant death or severe injury by head trauma. From what I remember of trauma victims (it's been a while) the head injured didn't make it to us in the hospital, but even the serious chest injured survived. I would guess it's the head trauma that kills most people in side impacts with SUVs.

    It is interesting that the IIHS side impact test differs so much from the NHTSA test in terms of torso and pelvic injury, if you compare the exact same car tested and ignore the head injury score. (look at the Accord without side airbags) Maybe it has to do with the size of the crash dummies? It's also odd that the rear passenger did well in the crash. The rear passenger would be subject to the same lateral force as the front passenger (although the intrusion may be less-but the picture showed damage all the way to the rear tire) I wonder if the IIHS test has been validated with any real world data? It's probably too new of a test, though. I wonder why so many cars do so poorly in this test? Maybe the insurance companies can use it to justify higher rates?

    I fully agree with what I think Craig said- it may be that you put an average sized male in the same situation and the results may be very different. This is the problem with laboratory testing- it doesn't necessarily correlate with the real world. For example, the GM minivans did very poorly in the IIHS offset crash test- yet real world data showed in accidents it had some of the lowest injury ratings.

    I would say that considering the safety engineering that went into this car, and how well it did with crash testing in Australia, that it is a safe car, and in fact far more than most of the other cars that have been tested. It did very well protecting the head, which is most important. It will be interesting to see how the NHTSA side impact testing goes. Until then, I will not overreact to the results of one test which contradicts other independent testing. It may also be that the wagon (and outback) do better in side impacts then the sedan (as the 04 legacy wagon did better than the sedan in the NHTSA side impact testing)

    If you want to talk death trap, go look at a Ford Pinto.

    tom
  • Yeah, melodramatic is a good word for this instance. On the bright side, if everyone freaks out about the '05s it should make it easier to haggle the dealer down.

    We have on test in Aus. that shows an exceptional side impact rating and a US test showing an Average rating.
  • lmacmillmacmil Posts: 1,758
    This whole crash test hysteria blows my mind. What is the probability of any of us being involved in an accident, much less an accident that duplicates the exact conditions of these tests? Probably about the same as being struck by lightning or winning the Powerball lottery. If you are that worried about being injured in a car accident, buy a Hummer or an Excursion.

    I am going to unsubscribe to this group until you get back to talking about the Suburu driving/ownership experience instead of the hypothetical Suburu crash experience.
  • roopower2roopower2 Posts: 13
    While pondering the side-crash test results on the new Legacy (which I hope to purchase in 2-3 years), I decided to drive my 1981 Subaru BRAT to work today. This vehicle has no internal door braces which were required in 1981 for cars, but not trucks. Airbags-nope. Do I worry? Not really, at least it will be quick!
    John
  • snowbirdsnowbird Posts: 120
    Interesting to note that, according to Consumer Reports, the 2004 Legacy/Outback had a IIHS test rating of "Good"- top ranking. Yet, this year, the same agency rated the 2005 Legacy "Marginal" - 2 ranks down and just above "Poor". So what gives? I thought we were supposed to get a better and safer car in 2005. No?! (Confounded) Snowbird.
  • roopower2roopower2 Posts: 13
    While pondering the side-crash test results on the new Legacy (which I hope to purchase in 2-3 years), I decided to drive my 1981 Subaru BRAT to work today. This vehicle has no internal door braces which were required in 1981 for cars, but not trucks. Airbags-nope. Do I worry? Not really, at least it will be quick!
    John
  • lark6lark6 Posts: 2,565
    So I did the right thing by sticking with the Forester? ;-)
  • Only if you like the Forester better then the Legacy. The Legacy is NOT an unsafe car by any stretch. Looking at the numbers and ratings, most of the cars tested by IIHS this round got "Poor" scores across the board. One telling sign, the deformation from the impact does not actually "hit" the driver in the legacy as it does on most of the cars in that test. My plans to buy an '05 GT have not changed at all. If anything this panic over above average, for this test, scores will just lower demand and make it easier for me to get a better deal :)
  • robr2robr2 BostonPosts: 8,863
    One major flaw in all the tests is that they do not take enough data points to have a statistically valid sample. It may be that the variation amongst many cars of the same type is low, but I have not seen any indication that they test enough cars to even come to that conclusion. I'd be very curious to know how much the results change from car to car if, say, 25 Legacys were tested.

    First, who is going to fund the actual crashing of 25 vehicles?? In reality, the variance in results should be extremely low - these are highly engineered, consistently manufactured products with very tight tolerances. The variability from vehicle to vehicle should be nil or close to it. Now it's possibile that IIHS got a bad unit twice, but I doubt it.

    It's hard to make valid conclusions based on very limited data when trying to predict the performance of any machine under a wide range of conditions.

    Again who's going to pay to test every possible scenario? The IIHS perfoms a consistent test in a consistent environment for everyone. No variability delivers results that can be compared.

    In reality, the Legacy should still be an extremely safe automobile for the majority of the population. Keep in mind that the IIHS is funded by INSURANCE COMPANIES. They have a vested interest in determining what they consider a safe automobile.

    Lastly, I was under the impression that many manufacturers are now conducting virtual crash tests using computer simulation during the design phase. These results should not be a total suprise anymore or there's some software tweaking to be done.
  • lark6lark6 Posts: 2,565
    Only if you like the Forester better then the Legacy. The Legacy is NOT an unsafe car by any stretch.

    Oh I know, I probably should've used a ;-) emoticon in the last message. (There, fixed it. ;-))
Sign In or Register to comment.