Collision with Marketing Traps: A Sturdy Car Does Not Equal Safety | A Letter from an Engineer

When crash tests become a marketing tool, how can consumers identify vehicle safety?

Li Auto showcased a video of a head - on collision between a 2.6 - ton SUV and an 8 - ton truck during its new car launch event. In the video, the two vehicles collided at a relative speed of 100 km/h. Subsequently, the truck's wheels bounced up, and its cab "nodded and hit the ground," while the A, B, and C pillars of the Li i8 remained undeformed, all airbags deployed, and the battery did not catch fire.

This "safety show" that defies physical common sense quickly sparked doubts and dissatisfaction among consumers. In fact, using crash tests to prove safety has become a standard practice at new energy vehicle manufacturers' press conferences.

During the launch of the NIO LeDao L90, a scene was presented where the vehicle was sandwiched between two trucks at high speed in rainy conditions, and the vehicle's body frame remained intact. BYD's Sea Lion 06 conducted a static roof pressure test, claiming that it could withstand approximately 9.9 tons of pressure, equivalent to the weight of two elephants.

The common characteristics of these tests are strong visual impact, special test conditions, and results that go beyond common sense. Automobile manufacturers attempt to use "explosive" test results to obscure consumers' understanding of the essence of safety: the real safety of a vehicle lies in protecting passengers and long - term reliability, rather than "special - effect performances" in specially designed laboratory scenarios.

Behind the marketing chaos, consumers' doubts about vehicle crash safety have never dissipated.

Currently, the main domestic institutions that provide vehicle crash tests include CATARC's C - NCAP, CIASI's C - IASI, as well as commercial platforms such as China Automotive Engineering Research Institute and Dongchedi. However, the rating standards of these institutions have not been widely recognized. Questions such as how these institutions obtain test vehicles and whether automobile manufacturers can manipulate the vehicles submitted for testing have long remained unanswered.

Simo, an engineer engaged in vehicle crash safety work, told 36Kr that since it is a common industry view that the standards of Chinese testing institutions are lower than international ones and the scoring methods are unreasonable, engineers involved in crash safety rarely pay attention to these crash test results.

"Even if a vehicle gets a five - star rating, it doesn't directly indicate that the vehicle has excellent safety. What is truly related to safety is the in - house crash tests of the automaker. The automaker's standards must be higher than those of any third - party."

Amidst the marketing scripts and doubts about evaluation data, consumers need professional interpretations to penetrate the data fog. What aspects of vehicle crash safety are really worth paying attention to? Simo shared more with 36Kr.

The following is a conversation between 36Kr and vehicle safety engineer Simo, slightly adapted without changing the original meaning:

36Kr: Which crash testing institutions are highly recognized globally? What are the characteristics of their evaluation systems?

Simo: Globally, the Euro - NCAP in the EU and the NHTSA - NCAP in the US are relatively well - recognized. Chinese cars are rarely exported to the US, so there isn't much data on the N - NCAP standard, and there's no need to pay attention to it.

The Euro - NCAP is known for its strictness. First, it does not accept vehicles submitted for testing; it only conducts random inspections. Second, the standards for many tests are the highest in the world. For example, the speed requirements for some crash tests are about 10 km/h higher than those in China. Many current tests and test standards originated from the Euro - NCAP and have been promoted globally.

Moreover, this is not just a reference standard but more like an access standard. If a vehicle fails to achieve a three - star rating in the Euro - NCAP, it is very likely to be taken off the market.

For Chinese cars to obtain Euro - NCAP scores, they must be sold in Europe. There is some data available for reference, but not much.

36Kr: What about Chinese crash evaluation institutions? Which ones are reliable?

Simo: The well - known ones are CATARC's C - NCAP and CIASI's C - IASI. Nowadays, platforms like Dongchedi also conduct crash tests.

Let me first talk about the positive aspects. Whether it's C - NCAP, CIASI, or other platforms, Chinese crash evaluation institutions are ahead of other countries in the field of electric vehicles, especially in battery crash safety.

However, domestic evaluation institutions each have their own drawbacks. Dongchedi's test items and standards are relatively in line with current hot topics. But since it is a commercial company, there are bound to be commercial motives.

As for C - NCAP, its test items are generally in line with those of the Euro - NCAP, but the standards are slightly lower. If a vehicle passes a test with a non - threatening speed, there's no reason to celebrate. If it can't even pass this, then the vehicle shouldn't be launched on the market.

Moreover, the scoring rules of C - NCAP are different. First, to promote the popularization of vehicle configurations, C - NCAP has many bonus items. For example, if a vehicle is equipped with a rear - seat unbuckled seatbelt reminder function, it gets an extra 1.5 points. It's not that cheap to improve the score by 1.5 points in the occupant crash test.

For a new vehicle to get a five - star rating in the Euro - NCAP, the worst - performing item must at least achieve a four - star rating. But C - NCAP has no such requirement.

Finally, consider that C - NCAP accepts vehicles submitted by manufacturers for testing, and the annual five - star rating rate exceeds 70%. The purpose of consumers looking at evaluations is not to see how good all the vehicles are but to identify which ones are good and which ones are not. Since all the tested vehicles are rated as good, the credibility of this institution is questionable.

CIASI is more reliable because of its independent vehicle - purchasing method. However, its standards are different. A high - level rating from CIASI doesn't necessarily mean high - level safety; it may be an assessment of the economic cost of vehicle repairs. For a car like an Audi, where replacing a headlight costs 10,000 yuan, its score in some CIASI test items won't be very high.

So, when looking at evaluation institutions, one should first look at the standards to understand what the scores actually represent, rather than assuming that a high score means the vehicle is the safest.

36Kr: So, what should a credible crash test be like?

Simo: I can talk about how automakers conduct crash tests.

First, everyone should understand that the fundamental purpose of conducting crash tests is to improve vehicle safety, minimize the casualties of vehicle occupants during a crash, and there are also some indicators to protect pedestrians.

So, when looking at a crash test, the core is not to see whether the front or rear of the vehicle is damaged. The body - in - white is just the first line of defense for safety. If the body - in - white doesn't have sufficient strength, the vehicle won't be launched on the market.

The real data of a crash test should focus on the injury situation of the dummy models inside the vehicle. This is why many automakers like to promote how expensive the dummies are and how much they've invested in purchasing them.

Secondly, within automakers, the scenarios and intensities of crash tests are more diverse than those of testing institutions. For example, people may think that as long as the A - pillar is strong enough, the crash result will definitely be safe. That's not the case. Differences in dummy models, the positions of the steering wheel and airbags, the angle, speed, and force of seatbelt tightening, etc., even a one - degree difference in these data can lead to different results.

36Kr: So, do you think the test results of these institutions have little reference value?

Simo: For institutions like the Euro - NCAP and C - NCAP, their test items are single and fixed scenarios, and the standards are public. Passing these tests doesn't directly mean the vehicle is excellent. However, a vehicle that can't even pass the C - NCAP test probably shouldn't be launched on the market.

36Kr: Will automakers make targeted optimizations for these tests?

Simo: The types and standards of in - house crash tests in most automakers are higher than national standards and those of various evaluation institutions. To be honest, our team hardly pays attention to domestic institutions like C - NCAP. We've conducted all the tests they do, and we've also done some that they haven't. There's no need to pay attention.

Of course, the certification team will pay attention to some regional access standards. If our cars are to be sold in Europe, we definitely need to pay attention to European crash regulations and pass their tests.

These non - access evaluation standards have little reference value for automakers' R & D. If an automaker focuses most of its work on meeting the standards of these institutions' test items, it's hard to evaluate the quality of its vehicles.

36Kr: What's the difference in crash performance between high - end and low - end models of the same vehicle?

Simo: There are often differences in materials between high - end and low - end models of many vehicles. For example, high - end models may use aluminum alloy, while low - end models use steel plates. High - end models may have more airbags. Some unofficial tests may play tricks by comparing a high - end model with a low - end one. So, generally speaking, the data from institutions that purchase low - end models are more reliable.

36Kr: Is the experiment of a passenger car crashing into a truck meaningful?

Simo: I think when ordinary consumers look at a crash test, they should first focus on the purpose of the test.

Crash tests are mainly divided into two categories. One is the body structure strength test, which is usually conducted repeatedly during the body - in - white stage without waiting for the whole vehicle to be assembled. The other is the crash safety test, which is the core of the whole - vehicle crash test. The focus is to understand the injury data of dummy models under various crash conditions.

The experiment of a passenger car crashing into a truck can at most prove that there are no major problems with the body structure, but it doesn't mean that the vehicle occupants won't be injured. And colliding head - on with a truck is not the best or most effective way to prove body structure strength.

Moreover, crash tests generally need to be conducted in a professional venue, and professional equipment is required to accurately control acceleration. If people drive the cars to collide, it's questionable. How can you keep the speeds consistent? How can you control the acceleration? If these factors can't be controlled, how can the results be reliable?

36Kr: The public generally believes that a "hard" car means safety. But in your opinion, a "hard" car doesn't necessarily equal safety?

Simo: Vehicle safety is a broad topic. In fact, the R & D of every component is related to safety. Some established enterprises even have their own independent safety systems with their own standards for every link from R & D to testing.

The perception that a "hard" car means safety is an impression left on the public by the past.

In the past, the materials and manufacturing processes were not as good as they are now. It was challenging to build a body - in - white with sufficient structural strength and torsional stiffness. Also, the technology of simulation dummies was not well - developed 20 years ago. So, the crash regulations at that time mainly focused on whether the body structure was intact and whether the doors could be opened smoothly.

However, now the crash regulations have changed. The current focus is not on whether the A, B, and C pillars are intact but on the injury situation of the dummy models inside the vehicle after a crash. And the dummy models have gradually expanded from data based on European white men to include female, child, and elderly dummies.

For example, if a car collides with a tree on the side of the road at a speed of 60 km/h in an offset collision. Even if the body strength is excellent and the A, B, and C pillars remain intact, the vehicle occupants will definitely be subjected to forward and backward impacts. If the seatbelt doesn't tighten and can't secure the passengers, the passengers' heads may repeatedly hit between the window and the seat, resulting in a concussion and facing the risk of injury or even death.

So, you'll notice that many automakers now emphasize that the seatbelt will tighten during an accident and that airbags are installed in certain corners. These facilities that prevent internal impacts are the key factors affecting the injury situation of vehicle occupants at present.

Of course, all of this is based on the premise that the body - in - white is excellent. If a sedan gets under the tail of a truck and the A - pillar is completely damaged, the airbags will be useless. However, the manufacturing difficulty of the body - in - white has been greatly reduced. The competition among automakers is no longer about maintaining a minimum standard but about raising the upper limit.

So, the safety index of vehicle occupants is actually closer to the core of vehicle safety.

36Kr: Which is more difficult to improve, the strength of the body - in - white or the prevention of internal impacts on vehicle occupants?

Simo: Different enterprises face different levels of difficulty. Some enterprises have had their own crash centers since 1980 and have been conducting crash tests for more than 40 years. For them, improving the body - in - white strength is not that difficult.

Currently, the main difficulty in the industry lies in preventing internal impacts on vehicle occupants. I know that some enterprises have specifically formed a team of more than 20 people to be responsible for the design of internal impact prevention in crash safety.

This is a meticulous and complex task.

First, there's the issue of dummy models. What parameters of dummy models should be used? Currently, the regulations only specify one type. If an enterprise wants to introduce more dummy data, such as for the elderly and women, it requires the enterprise's decision - making.

Enterprises with a strong safety awareness may test the crash results of a dummy in different positions in the seat and with different steering wheel positions to ensure that their customers are well - protected in any scenario.

Then there's the design of crash scenarios. The regulations require less than 10 scenarios, but enterprises can't limit themselves to these. At least 10 - 20 scenarios are needed. Enterprises with sufficient funds and experience may conduct 80 - 90 scenarios. Considering all these factors, it's not excessive to assign a 20 - person team to handle this work.

36Kr: So, does safety depend on a sufficient number of crash tests?

Simo: To some extent, yes. If you have money, you can conduct real - vehicle crash tests; if not, you can do simulations. Whether it's improving the body - in - white strength or preventing internal impacts, a large amount of cost is required. Our job is to optimize vehicle design based on these crash test results.

36Kr: For consumers who don't know much about cars, what items do you recommend they pay attention to?

Simo: The scenarios closest to daily accidents, such as high - speed offset collisions, are worth paying attention to. In addition to checking whether the doors can open and the airbags can deploy as shown in some promotional videos, the most important thing is to look at the injury data of the dummy models. Especially now, consumers may be more concerned about data for female family members and children.

Most enterprises don't disclose their in - house test results. If consumers want to know, they can visit the official websites of CATARC and CIASI. Don't just look at the rating results; instead, watch the test videos and look at the specific injury data released by the institutions.