Abstract:
American physicist Richard Feynman wrote in the Challenger accident investigation report: "For a technology to be successful, reality must take precedence over public relations, because nature cannot be deceived." The brake pedal assembly may be one of the most mature parts of the car. Recently, it has become a footnote to this sentence.
Article Zeng Xing Zhao Yu Li Anqi Guo Ruichan Li Zinan
Editor丨Gong Fangyi
In the past two days, "LatePost" has communicated with a number of automotive R&D engineers and scholars. Their majors involve brake R&D, materials and failure analysis. Although the starting points of the analysis were different, their first judgment was almost the same: no matter how the driver stepped on it, the brake pedal bracket should not be broken.
“After all, it shouldn’t have broken.” A brake engineer who once worked in foreign Tier 1 repeatedly told LatePost. He has been engaged in brake research and development for nearly ten years, and said that during the development process, companies will also ask testers to depress the brakes as hard as they can, because users cannot be required to control their pedaling intensity in an emergency.
A brake engineer from an independent car company was also surprised. He said that when they develop the brake pedal assembly, they will check the static strength, impact strength and durability performance, and will continue to increase the load to find the boundary of its fracture; when doing internal brake testing, they will also learn from the media to step on it. In his view, parts that have undergone multiple rounds of complete verification should not be exposed to such problems until the vehicle is delivered.
"Car consumers who have experienced refrigerators, color TVs, large sofas, tents, sleeping bags and mahjong tables should return to rationality and examine what should be valued most." Another veteran with more than 20 years in the industry said.
People close to the incident said the fracture of the Zunjie V800's pedal bracket was accidentally discovered during routine testing. The test team subsequently repeated the same scenario two more times, and the pedal brackets of all three vehicles were broken. On the evening of October 8, Zunjie Auto said that its braking system has completed the full-process development and verification in accordance with the requirements of national standards GB 21670-2025, GB 7258-2026 and industry standard QC/T 788-2018.
Researchers engaged in thermoplastic damage behavior and material mechanics testing at Tsinghua University focus on the coordination of materials and structures. He believes that the glass fiber reinforced plastic used in the Zunjie V800 can meet the usage requirements of the pedal bracket, but the performance of the material will be different when pedaling quickly and slowly applying force, so the design needs to take these differences into account, as well as fluctuations caused by temperature, material batches and processing, and leave sufficient safety margins.
Relevant national mandatory standards only require a pedal force of 500 N to brake the car, while industry recommended standards require the bracket to withstand 1500 N. "LatePost" learned that the requirements within companies are often much higher: SAIC-GM's design starting point is 2,700 N, the damage load of a traditional car company is generally not less than 3,000 N, and the design strength of a leading new force is also higher than 3,000 N.
"I don't care about the specific value. This is a matter of research and development. I care about the impact of this matter. Risk analysis has three dimensions, severity, frequency of occurrence, and discoverability. The severity is close to infinity." A purchaser from a joint venture brand said. "If the national standard cannot be covered, then the national standard must be changed, and the tester will be called over to calibrate the new standard on site."
Yang Zhenguo, a professor at Fudan University and who has studied failure analysis for more than 40 years, emphasized that confirming the root cause requires obtaining fracture samples. Materials, construction, processing, assembly and operation may affect results. Videos and photos alone cannot determine the priority of these factors, let alone a substitute for physical inspection and testing.
The interview also gradually touches on how cars are developed today. Several engineers mentioned that the model development cycle is shortening and cost requirements are becoming more and more detailed. Reusing mature parts, carrying out work in parallel, and transporting cars to the southern hemisphere for winter testing can speed up development, but it also brings more uncertainties.
In 2010, when Akio Toyoda went to the U.S. Congress to explain the recall crisis, he admitted that Toyota's traditional ranking of safety, quality and sales had become chaotic during rapid expansion. "We're pursuing growth faster than we can develop people and grow the organization," he said.
The following is a conversation between LatePost and 9 automotive engineers, academics, and industry professionals. Some spoke on condition of anonymity.
I have never encountered this in the past ten years, "In the final analysis, it should not be discontinued"
Later
: As a practitioner, how do you understand Che Di’s test?
Brake engineer of independent car company
: Rarely. The pedal has been around since the middle of the last century and is very mature. But the video of Understanding Car Emperor is more rigorous. The same test was done on multiple cars, and different cars were used for comparison. The test data was displayed in one shot. The pressure-sensitive force sensor he used is also a common tool in the industry.Foreign-invested Tier 1 brake system R&D engineer: I have been working on brakes for nearly 10 years. The plastic piece on the front of the pedal broke when someone stepped hard and the foot became crooked. But I have never heard of the location where the shaft broke like this.
This is a purely hardware design problem. If three cars in a row are broken, there must be something wrong, there is no doubt about it. Looking at the video, it seems that the people who tested the first two cars used their heels to step on it. Normally, they use the balls of their feet to step on it. This is not a problem either. No matter how you step on it, as long as the driver can depress it, it's a problem with the brake pedal assembly itself.
Later
: Based on the fracture process shown in the video, what might be the problem with the part design?
Researcher on Materials Mechanics at Tsinghua University
: Looking at the fracture location from the video and pictures, there should still be a problem with its structural design. The 100-kilometer emergency braking operation is a very short and very fast operation. When designing, it may ignore some characteristics of this material under dynamic loads.Brake engineer of an independent car company: Its pedal form is not special, but looking at the position of the pedal shaft in the video, the size of the base around it is relatively small. After the pedal is stressed, it is actually a seesaw. When you step on the pedal, the rotating shaft will tilt upward, so the two bases pressing the rotating shaft carry the upward force and are generally not made so thin. We really haven't seen anything this thin.
Later
: When the brake pedal bracket of the third Zunjie V800 tested by the car expert was broken, the sensor reading was 1612 N. How do national standards stipulate the endurance threshold of pedals?
New force brake engineer
: The national standard has a mandatory value of 500 N (this force is equivalent to a 51 kg person standing still on the scale), which is the requirement for braking portability: 500 N is required to reach the maximum braking force or maximum deceleration, allowing drivers with little strength to brake the car. There is also a recommended index, 1500 ± 20 N, which is the requirement for the mechanical strength of the brake pedal. One is the braking performance requirement, and the other is the upper limit recommendation of the mechanical structure.Brake engineer of an independent car company: Under longitudinal load, the QC/T (automotive industry recommended standard) standard is 1500N ± 20N; when we set the group standard before, it was based on 2000N. Neither of these two standards are mandatory national standards. GB 7258 (a national standard for motor vehicles with a wider scope) only states that the pedals must be structurally stable and cannot be damaged. The 500 N mentioned in GB 21670 (a national standard specifically for braking systems) it quotes refers to the force value at which the vehicle can be braked. Therefore, from the perspective of national standards, it is not wrong to say that Zunjie meets the national standards. It's just another thing if something goes wrong when someone steps on it like this.
Foreign-invested Tier 1 brake system R&D engineer: 500 N is for the brake controller, that is, 500 N can output the maximum braking force. But what is guaranteed on the pedal is definitely not 500 N. What is guaranteed is that as hard as a person can step on it, it will continue to withstand it. This is the most basic: the accelerator can be cut off, but the brake cannot be cut off.
Later
: Can a normal person pedal with a pedal force of 1612 N?
New force brake engineer
: This force cannot be achieved by standing still and pedaling. During the braking process, the weight caused by the deceleration of the body is also transferred to the pedal, which is both active force and body weight. Therefore, this intensity is not outrageous at all. In our own testing, it may be exaggerated and there will be more rough testing methods. There are also errors in the data collected by different devices, but overall it is quite reasonable.Foreign Tier 1 brake system R&D engineer: The data I can give is that a normal test driver, who is relatively strong and burly, can pedal 1800-2000 N. With a pedal force of 1800-2000 N, a normal vehicle will definitely be stepped on more than once during the development process and vehicle road test. We've installed pedal force sensors before, and that's about the same number. In an emergency, people's adrenaline kicks in and they step harder.
As long as you exceed 500 N, the final braking force the system gives you will be the same. But you can’t ask customers to step on 500 N when they encounter an emergency. This is unrealistic. Normally, the brakes should not break as much as you can.
Later
: How much extra allowance do car companies generally make for the strength of the brake pedal?
Brake engineer of independent car company
: We will definitely do the conventional 2000 N standard, and we will also deliberately load it up to the point where it may break, and determine what the load is when it breaks. It’s hard to tell the specific value, but you can understand that it is generally not less than 3000 N.Supplier controller development engineer: According to our experience, the human foot can step out at most about 2000 N. Multiply this number by the leverage ratio: 2000 N at the far end, which gives a leverage ratio of 4, which corresponds to 8000 N at the near end of the axis. The destructive power must be at least 8000 N, and with design redundancy, it can reach 30%, 50% or even 100% according to industry practice, which is more than 10000 N. You said it broke at 1600 N, that is certainly unreasonable.
Later
: Although the Zhijie and Wenjie models compared in the video of Understanding Cars also use plastic brackets, they are not broken.
Researcher on Materials Mechanics at Tsinghua University
: I guess Zunjie’s brake pedal bracket design doesn’t have enough redundancy. When comparing the two cars that I understand, there are enough materials in that area, and there are more materials to bear the same load. The possibility of crossing the critical point and entering yield is very small. The Zunjie V800 brake pedal bracket has very little material. From the video, there are only a few hollow structures with reinforced ribs. Just like the total load is 1500 N, whether it is carried by 10 people or carried by 100 people, the force borne by each person is different.I don’t know what they considered in the design. It may be to reduce weight and save materials. But there is a lot of weight that can be reduced in other places on the car, so there is no need to lose weight in such a critical position.
Later
: What is the connection between the damage to the pedal structure and the performance of the braking system itself?
Foreign Tier 1 brake system R&D engineer
: It is the structural point above and has nothing to do with the brake controller. The brake controller makes it clear that over 500 N will provide maximum braking force. It was the lever structure and the shaft on it. It couldn't withstand such a large force and broke. It's purely a hardware problem. It has nothing to do with how powerful the braking system is, or whether it's electronic or hydraulic. The problem is the brake pedal assembly, not the braking system assembly.Later
: Will a recall be triggered after this happens?
Industry people close to regulatory authorities
: Car recalls are initiated by car companies. If they don’t take the initiative to recall, there is basically nothing they can do. Car companies can solve the problem with free upgrades, and the government also agrees. As long as the problem is solved, it doesn't matter whether it is recalled.Later
: If there is a major safety hazard, as long as the car companies do not take the initiative to recall it, is there nothing the government can do?
Industry people close to regulatory authorities
: This can be discussed. Recall is not a management tool, but a voluntary mechanism. If car companies do not take the initiative to recall, the regulatory authorities may not issue them 3C certificates. Every car must have a 3C certificate, without which it cannot have a license plate.Later
: If a model has been delivered to relatively few vehicles, can free upgrades actually play a role in recalling it?
Industry people close to regulatory authorities
: Most recalls by Chinese car companies are implemented through free upgrades. Looking at the past 20 years, few Chinese car companies have taken the initiative to recall vehicles. This is a matter of concept: for foreign car companies, recalls are positive public relations. The more recalls, the more responsible the brand is. For Chinese car companies, recalls are a shameful thing.Design, materials and manufacturing, each link may leave hidden dangers
Design: The safety margin of the bracket needs to cover the force exerted by emergency pedaling
Later
: It’s the same brake pedal bracket. Why do different models have different structures?
New force brake engineer
: The pedals are actually customized components. Each model is affected by the vehicle layout, front cabin structure, and space design. The pedals will be different. In addition, pedal design is also affected by the technical solutions of different suppliers. Each supplier's testing standards and methods will be somewhat different. Some suppliers have more redundancy in their designs, and some have less.Brake engineer of an independent car company: Judging from our experience, Zunjie’s pedals are different from those of the other two. It may be that MPVs are relatively heavy, so in order to reduce weight, they have made such a design.
Later
: How to determine the safety margin of the base during design?
Brake engineer of independent car company
: Different companies have different design requirements for pedals, and there is no unified standard in the industry. Whether it can pass or not requires preliminary simulation through computer simulation during the design stage, such as finite element simulation. Through simulation, we can see whether the deformation of the pedal can be controlled within the target range. If the simulation is passed, there is a high probability that it will not be a problem. If the deformation caused by the simulation is relatively large, the plan will generally need to be revised.However, there is a problem with finite element simulation. Its transient dynamic load (a load whose magnitude, direction or point of action changes rapidly with time) simulation is not accurate and can generally only be used for reference. First, there is an angle when people step on it, and they do not step on it steadily; second, there is an instantaneous acceleration when people step on it, and there is an instantaneous impact. That is to say, the actual load is a force that rushes up first, and then returns to what we generally call 2000 N, 1600 N, or 1500 N. From the perspective of the force, everyone imagines that when stepping on the pedal, they put their feet up and down slowly. In fact, many people use the inertia of their body to step down.
Later
: What aspects does the redundancy of brake-by-wire generally cover? Why are mechanical inputs like pedals usually not included in redundancy?
Wu Xiaodong, associate professor at Shanghai Jiao Tong University Intelligent Vehicle Research Institute
：EHB (Electronic Hydraulic Braking System) The body design redundancy of brake-by-wire generally includes controller and power supply redundancy, hydraulic execution circuit redundancy, and signal sensing redundancy. In addition to the main body, there are also external assistance such as motor braking and EPB after failure. The current ISO26262 functional safety design for automotive electronics mainly targets random failures of electronic and electrical hardware. Mechanical fracture of the pedal bracket is classified as a mechanical structural failure. Mechanical systems generally only verify reliability through strength, durability, and fatigue.Later
: After the pedal bracket breaks, is there any way for the braking system to identify "pedal failure" and distinguish it from "the driver did not apply the brakes"? If it can be identified, should the system actively degrade braking?
Wu Xiaodong, associate professor at Shanghai Jiao Tong University Intelligent Vehicle Research Institute
: The pedal is only mechanically broken. If the system circuit is intact, the system cannot definitely identify the pedal failure, so it cannot unconditionally initiate active degradation braking. Of course, after this incident, it is not ruled out that car manufacturers will consider designing identification systems to downgrade, but this will increase costs.Later
: Brake-by-wire is moving from electro-hydraulic braking (EHB) to electro-mechanical braking (EMB), with fewer and fewer physical connections between the pedals and wheels. Under this trend, should the reliability requirements for the input end of the pedal also be increased?
Wu Xiaodong, associate professor at Shanghai Jiao Tong University Intelligent Vehicle Research Institute
: The traditional hydraulic brake pedal performs two functions at the same time: braking intention acquisition and mechanical force transmission. However, the brake-by-wire pedal has only one function: braking intent collection. Therefore, the reliability of the pedal as the source of all braking inputs is particularly critical. Subsequent product development must increase the design margin of the pedal mechanical body, and fully incorporate mechanical failures such as pedal bracket breakage into the vehicle hazard analysis, and no longer simply classify them as non-considerations.Material: Glass fiber reinforced plastic can meet the stiffness requirements, but it is also more afraid of sudden braking
Later
: What is the difference between adding glass fiber reinforced materials to polypropylene and traditional metal pedals?
Professor Yang Zhenguo of Fudan University
: The difference is huge. Car pedals have roughly gone through three generations: the first generation is made of metal, but the disadvantage is that it is too heavy; the second generation is made of continuous glass fiber thermosetting resin fiberglass, which reduces the weight, but the material is brittle and cannot be recycled; the third generation is glass fiber reinforced PP (polypropylene), which can be recycled and reused, and the weight is reduced. PP itself can be divided into isotactic, syndiotactic and atactic polypropylene, and there may even be a small amount of PP recycled materials with different properties. Nowadays, in order to reduce weight, medium-length glass fiber reinforced isotactic polypropylene composites are mainly used.Later
: Is this material enough to make a brake pedal bracket?
Researcher on Materials Mechanics at Tsinghua University
: If the structural design is reasonable, there is no problem for this material to meet the requirements of the pedal. Similar models use similar materials and there will be no problems.Later
: The national standard requires pedal materials to have properties comparable to those of metal. How do you understand "quite"?
Professor Yang Zhenguo of Fudan University
: That is, the key performance is basically the same. From a material perspective, the “four degrees” are mainly considered: strength, stiffness, toughness and hardness.Strength is the stress endured by the cross-section of a component, such as tensile strength and yield strength; stiffness is the deformation of the component under the action of external forces, such as deflection, bending, and twisting; toughness is the ability of the material to absorb energy and plastically deform before fracture, which includes resistance to crack fatigue expansion and resistance to fracture under impact loads; hardness is the wear resistance of the material against local deformation of the surface, as pedals are often stepped on.
Later
: Compared with braking slowly, plastic parts are more likely to break if you slam on the brakes suddenly, is that true?
Researcher on Materials Mechanics at Tsinghua University
: It’s okay to roughly understand it this way. The mechanical properties of plastics and metals are quite different. The current national and industry standards mainly focus on the metal brake pedal bracket structure, and other materials are only "available for reference".Take the metal pedal bracket as an example. The industry standard requires that the pedal be stepped on with a force of 1500 N and held for a certain period of time. After being released, the permanent deformation cannot exceed 5 mm. In fact, it requires elastic deformation and recovery. The metal bracket meets this requirement, and it will probably be fine if you step on it hard. Because of the mechanical behavior of metals such as steel and aluminum in the elastic stage, slow loading and fast loading are basically the same.
However, polymer plastics are softer under slow loading and harder under impact, and their elastic modulus becomes larger. This feature sometimes causes problems. For example, if a plastic is slowly loaded to 5 mm, it is still within the elastic range and can recover from deformation; but under impact, its elastic range may be only 3 mm and not 5 mm at all.
Later
: The deformation corresponding to the elastic limit becomes smaller.
Researcher on Materials Mechanics at Tsinghua University
:right. The original factory design may have been designed for a slow loading of 1500 N, which meets industry standards, but the dynamics may not be measured. If the design happens to be stuck at the critical point, it may exceed the yield limit at around 1500 N when stepped hard. The addition of glass fiber to this material ostensibly increases stiffness, but also introduces many interfaces and defects. Once the yield limit is exceeded, it will quickly become damaged, especially under impact. This damage can accumulate quickly under multiple loads.Later
: The car-savvy emperor did the test this time by slamming on the brakes with all his strength when the car was in relatively new condition. After picking up the car, ordinary users may press the brake lightly for the first dozens of times. If you know how to be a car expert and step on it lightly for dozens of times and then step on it harder, will the brakes be less likely to break?
Researcher on Materials Mechanics at Tsinghua University
: It has nothing to do with this. As long as you step on it once, for example, in an emergency, you have to brake to the end within a few seconds. This pedal structure may leave hidden dangers after you step on it once. It may not be broken the first or second time, but the damage is already inside and will not disappear. If you encounter this situation once or twice later, the damage will continue to accumulate and accelerate, and it may be broken by stepping on it the third or fourth time.Later
: Looking at the fracture of the part, do you tend to think that it was caused by fatigue or too much force in an instant?
Researcher on Materials Mechanics at Tsinghua University
: It broke after just a few steps. It probably has little to do with fatigue.Later
: What is the effect of temperature on material properties?
Researcher on Materials Mechanics at Tsinghua University
: Very significant. Especially when cold, engineering plastics like PP can be very brittle. This test should have nothing to do with temperature, but if the structural design does not take this difference into account, the probability of problems will increase in very cold places.Later
: Ford recalled more than 38,000 vehicles in 2020, also because of broken brake pedal brackets. Its bracket material is polypropylene and is broken at the axis of rotation. A report issued by the National Highway Traffic Safety Administration at the time stated that the root cause was that "noise factors" during the manufacturing process were not adequately considered.
Tsinghua University materials mechanics researcher: The performance of non-metallic materials is highly discrete, and I understand the "noise factor" to be fluctuations in performance. If only intermediate values of material properties are used in the design, deviations occur in a certain batch of materials, or defects are introduced during the manufacturing of a certain product, problems will arise. Whether it is a century-old factory or a new factory, foreign or domestic, if the design is not fully considered, there will be similar problems.
Manufacturing: Material batch and processing differences will affect the actual strength of the part
Later
: It is said that there are traces of plastic welding near the fracture of the Zunjie V800 brake pedal bracket. Is it normal for such marks to appear on injection molded parts?
New force brake engineer
:normal. For example, if you take a relatively rough or low-cost plastic, you may see something like a demoulding line. These are normal processing marks. Some will polish off the demoulding line, but those with lower cost may not polish it. This process is also part of the design - the demoulding line and the location of the so-called seam all need to be designed. If there is such a discontinuous shape or structure, you need to consider where it should be designed to have the least impact on the overall piece. It is a bit like a packaging tab - design where it will break.Later
: Some people believe that the break of the brake pedal bracket of the Zunjie V800 is related to the welding line.
Researcher on Materials Mechanics at Tsinghua University
: It is possible. The material for the weld line is indeed a bit weaker. However, if the design is well considered, the welding line can be placed in a location where the stress is not so concentrated through mold design, which can be avoided. If the fracture is really near the welding line, it means that the structure is already thin, and there is a weak link in the thin place, which is even worse. However, from the photos of the fracture you sent, it seems that it is not on the welding line. It still depends on where the stress is most concentrated on the bracket.The supplier of this component is actually well-positioned to discover the problem. They may have only done a test of pedaling at a slow speed and checked that there was no problem. A fast-loading test must be designed to test whether the pedal has any problems under impact.
Late: What variables will affect the performance of the part during the injection molding process?
Professor Yang Zhenguo of Fudan University
: The focus is on the molding process. The variables of the injection process include injection temperature, injection pressure, injection speed, runner and gate configuration, and the heat preservation process after the injection is completed. For components made of fiber-reinforced plastics, the rheological properties and viscosity of the melt must also be considered, which will greatly affect the distribution, dispersion and orientation of the reinforced fibers.The weld line is related to the molding process. As long as the mold is used for injection molding, there must be welding lines. The key is where it appears, which depends on the design of the gate, runner and mold. The welding line can be adjusted through the processing process design.
Later
: Regarding the injection molding process, can we make any judgment based on the probability and triggering conditions of fracture just mentioned?
Researcher on Materials Mechanics at Tsinghua University
: This is difficult to judge, because it is not clear what technology Zunjie uses.Later
: Is the consistency of different batches high for PP plus glass fiber? Could it be that the materials used in these three cars are of low quality?
Researcher on Materials Mechanics at Tsinghua University
: It is possible, but it is difficult to draw a conclusion. The biggest problem with non-metallic materials is that the consistency is not that good, worse than that of metals, and the difference is very obvious. It still needs to be evaluated from the structural design. Material performance has a range, and batch consistency should be included in the understanding of the material performance range. With this understanding, the lower limit of material performance should be considered during design and the structure should be designed using the lower limit. Otherwise, the performance of the structure may be overestimated.Foreign Tier 1 brake system R&D engineer
: Manufacturing problems are also possible. For example, the overall strength of this batch of materials is not good, and there is a problem with the injection molding process of this batch. But judging from the information that can be seen now, their production consistency is quite good, and the three methods are all the same. If manufacturing consistency is good, it's a design issue.Later
: How much difference can the quality of parts between small batch trial production and large-scale mass production differ?
New force brake engineer
: Test parts and mass-produced parts are two concepts. Parts are divided into soft modules, hard modules and mass-produced parts. For small-scale production, such as 100 pieces, the molds are sophisticated, the speed is not urgent, and the workmanship is better; for mass production, tens of thousands of pieces, the manufacturing process and parts level may be different. I really don’t know much about plastics, but for metals, the small batch workshop and the mass production workshop are generally not the same place, and there must be differences in design and processing operations. Not sure if it's a design, testing or manufacturing issue.Foreign Tier 1 brake system R&D engineer
: If it is a manufacturing problem, the car that was subjected to DV (Design Verification, Design Verification), PV (Production Validation, Product Verification) and vehicle calibration in the early stage may be different from the subsequent mass-produced parts. On the production side, has the injection mold verification of plastic parts been done? Have the parts coming out of the mold been inspected for size? Is it consistent with the digital model built in the previous stage? All may have an impact.Supplier controller development engineer: The status of molded parts may not be the same as those produced continuously. Therefore, destructive testing must be done on parts after continuous mold production and parts assembled off the production line, so that they can be closer to the state of mass-produced parts.
The brake pedal has a simple structure, but the verification cannot be simplified
Later
: The pedal seems relatively simple, is it usually made by the supplier?
Brake engineer of independent car company
: Some companies have outsourced it, but we insist on doing it ourselves. Most of it is designed by ourselves.Later
: Can’t we use the developed pedals in new models?
Brake engineer of independent car company
: Theoretically yes, because the load on the pedals mainly comes from people. If the pedal has been tested before and is relatively mature, it can be reused, but only if the previous test has been accurate.Later
: From pedal design to testing, what are the specific steps in the entire process?
Brake engineer of independent car company
: If the pedal design is done in-house, the design engineer will have a special design guide. After understanding it, he can start the design and conduct internal review; then he will conduct process verification evaluation and simulation verification for the supply chain before producing trial production. After the production trial is produced, it must undergo single piece testing, bench testing, and complete vehicle testing.Later
: How to do different tests?
Brake engineer of independent car company
: Suppliers need to do single piece testing and bench testing. Our bench testing will also measure several dimensions: static strength, impact strength, damage, and durability. The whole test still takes a long time. For example, the durability test usually takes 1 million times, and one round of durability takes at least a week, but we actually do many rounds.So in our eyes, from development to the end, there should be no pedal breakage. Although it is a very simple thing, its standards, requirements and testing, at the part level, installed on the assembly, and installed on the actual vehicle, are not simple at all.
Supplier controller development engineer
: The bench test is to disassemble the entire braking system separately and put it on the test bench for destructive testing. It is to measure the damage according to the limit value of the extreme working condition, which is equivalent to measuring the destructive force. It must withstand at least 1.5 times or even twice the redundancy under normal extreme working conditions. Zunjie may not have done this test in a targeted manner, otherwise this test condition should be able to be identified, and it is impossible to go directly to mass production in this state.Later
: How long does it take to complete the overall testing of the braking system? How much does it cost?
New force brake engineer
: The testing of safety parts such as steering/braking basically runs through the entire vehicle test. The vehicle testing period used to be basically 36 months. Normally, it was two winters and one summer. Now it is generally 18 months. Some car companies may do one winter and one summer, or even no winter, because the vehicles can be shipped to New Zealand in the southern hemisphere for off-season winter testing. The total number of tests remains the same but the time does get shorter.As for the actual vehicle performance test of the OEM, the cost of the new model calibration and testing is in the millions, and the cost of the replacement model is about 500,000, not counting the vehicle itself. But this is just a performance test, there are various other tests such as durability.
Supplier controller development engineer
: The whole vehicle test mainly lies in its road test. Under normal circumstances, road tests need to run at low temperatures and low temperatures, and at high temperatures and high temperatures, for at least 6 months from winter to summer to verify braking performance. Those who make brakes usually go to the Northeast for low-temperature calibration, and in summer they also need to do high-temperature calibration on high-temperature roads, because the friction coefficients and adhesion coefficients are different under high/low temperatures, and the friction coefficients on ice, wet roads, and cement roads are different.But if we just talk about pedals, it won’t take that long. It is a very simple structural component, not as complex as an electromechanical and hydraulic brake system. Just one pedal, the whole set of tests will take at most two months. From design to mold opening to production, it still takes a little longer, three or four months.
Later
: What verification environments does a brake component usually go through from design to production?
Foreign Tier 1 brake system R&D engineer
: The division of labor in the automobile industry is very detailed. A part has R&D engineers and design engineers, and the supplier also has its own design, project, and verification personnel. There are test engineers on both the OEM and supplier sides.Like brake components, it also involves brake system verification, vehicle calibration and road testing. Suppliers will do their own verification, and OEMs also have their own test requirements, and they also need to test on different roads such as high adhesion and low adhesion. From a system perspective, its verification redundancy should be very strong.
Later
: Will car companies replicate the performance of parts provided by suppliers to see if they can really achieve it?
New force brake engineer
: Normally, we don’t try to reproduce the hardware performance. The supplier will issue a test report, which is equivalent to a backup guarantee to facilitate subsequent accountability. Because in-person testing by the OEM consumes a lot of resources, not only the brakes, but also the chassis and suspension parts have to be tested by the suppliers themselves. Basically, suppliers endorse themselves.There will be improvements during the testing process. Make changes while testing and repeat testing. The OEM also makes requirements, such as testing directly after structural changes and reporting after the test. The staff of the main engine plant will analyze the report. They will not just believe in one result. They will judge based on the data: how many millimeters of deformation is the number of cattle, and the fatigue or durability strength after how many times of impact load.
Brake engineer of independent car company
: We need to check during the testing process. This is called process capability and process capability. During development, we need to check the production process capabilities, which will involve consistency issues. Our company's process must not only manage the capabilities of new parts or samples brought for testing, but also the consistency, all of which must be managed.Later
: If the requirements of car companies are very high, how will suppliers balance technology and cost?
New force brake engineer
: In the end, a compromise ratio of performance and cost acceptable to both parties will be reached, but the supplier's profits will be very low. Competition is too fierce now, and all suppliers and OEMs are operating under high-intensity, low-profit conditions. If the cost of a single piece is not suppressed, the entire cost will not be suppressed, and the selling price and profit will not be attractive. Of course, no matter how you choose, if the complete vehicle has not been verified, the OEM has a problem; if the parts are designed incorrectly and there is a large-scale performance decline in mass production parts and trial assembly parts, the supplier has a problem.Building cars faster and faster has benefits and costs
Later
: It is said that building cars is getting faster and faster now. How specific is this "fast"?
Foreign Tier 1 brake system R&D engineer
: Almost ten years ago, it took 4-5 years to develop a new model from project establishment to mass production. Now it should take no more than two years to do a new project.Later
: What specific decisions will be affected by time and cost pressure in the end?
Car company brake engineer
: The pressure falls on supplier selection, structural design, materials and processing technology. Just because no problems are found in the component's own tests, it doesn't mean there will be no problems after it's installed on the vehicle and put into the entire system.Foreign Tier 1 brake system R&D engineer
: Cars at different price points will have differences from supplier selection to specific design. Now the cost has been very detailed. For suppliers who only make mechanical parts, the profit may be only about 3%, or even lower.In order to enter the system of a certain car company, some suppliers may do the first one or two projects at a loss. They need to enter the system first. The more cars it sells, the more it loses, so much so that some suppliers even want car companies to sell less.
Later
: What would engineers worry about?
Foreign Tier 1 brake system R&D engineer
: Being fast has the advantage of being able to respond to the market faster. But from an engineering perspective, it is still too fast. In the end, it is consumers and car company employees who need to make up for these lessons.Car company brake engineer: The materials and processing technology are very complex. If there is a slight change, the final performance may be much worse. Sometimes the supplier feels that they just made a small change and tested it, but when it is put on the complete vehicle, the result may not work.
Later
: Does making parts thinner, lighter or changing materials mean cutting corners?
Foreign Tier 1 brake system R&D engineer
: You can’t judge that way. For example, after electric vehicles recover a large proportion of kinetic energy, the frequency of use and durability of friction brakes will decrease, and the brake pads can be thinned accordingly.This can not only reduce the cost of the friction plate and caliper assembly, but also increase the structural strength by shortening the span of the two arms of the caliper. Therefore, if you don’t think it is unsafe when you see “thinning”, it depends on whether the usage requirements have changed and whether the design and verification have changed accordingly.
Researcher on Materials Mechanics at Tsinghua University
: Nowadays, many parts introduce new non-metallic materials in order to reduce their weight, but the characteristics of non-metallic materials are quite different from those of metals. Most non-metallic materials are sensitive to loading rates. Whether you pedal fast or slowly, the difference in response is different from metal. This difference needs to be taken into account when formulating standards to make them more complete and stricter.In addition, there is currently no appropriate standard for evaluating brake pedal brackets made of non-metallic materials. According to the original standards for evaluating metal stents, it may not be comprehensive enough. When this kind of problem occurs, there may be trouble. The material selection has changed, and the corresponding testing and inspection standards must also be sufficiently considered.
Brake engineer of new power car company
: If it has not been verified at the vehicle level, it is a problem with the OEM; if there is a problem with the design of the parts themselves, it is a problem with the supplier. After a failure occurs, the two parts must be investigated separately, and the responsibility cannot be assigned to one party.Later
: With so many people participating in the verification, why do problems still occur in the end?
Foreign Tier 1 brake system R&D engineer
: The existence of a process does not mean that there will be no problems in every link. Several foreign companies I am familiar with at least have to go through a lot of verification and procedures. But if something that shouldn't be broken ends up flowing out, it means there is a problem with one or several links. The specific link is unknown to the outside world.Later
: Now that the fracture location is clear, why can’t we just study the broken component?
Professor Yang Zhenguo of Fudan University
: The fracture location is the source of failure, usually at the point of maximum stress or where a material defect exists. We saw in the video that a component was broken. This is only the macroscopic manifestation of the failure and cannot be directly equated to the cause of the failure. Just talking about one point can easily lead to one-sided and misleading conclusions. Many failures are not caused by a single factor, but are the result of the interaction and coupling of multiple factors. At present, there is a lack of on-site information and micro analysis, and it is not possible to prioritize various possible causes.I think the current analysis on the Internet is all based on the scope of one's own knowledge and ability to explain problems, which is not enough. Failure analysis is a comprehensive analysis across multiple disciplines. Any failure must consider the full life cycle safety assessment, including 11 factors such as design, materials, manufacturing, installation, inspection, operation, maintenance, environment, transportation, storage and management. As far as the accidental breakage of this bracket is concerned, related factors such as structural design, product quality, molding processing, assembly process, operating methods and environmental conditions need to be considered.
Later
: From seeing a component break to finally finding the root cause, what are the steps for a complete failure analysis?
Professor Yang Zhenguo of Fudan University
: I summarize the failure analysis system into "five elements": failure mode, failure form, failure defect, failure mechanism and failure cause. They can be mutually verified through the evidence chain, thereby ensuring that the analysis conclusions drawn are credible and reliable.Specifically, the failure mode is the appearance of the component after failure; that is, the macroscopic morphological characteristics of the failure; the failure mode is the specific location and damage characteristics of the component after failure. Also known as failure characteristics; failure defects are the original defects or secondary defects that cause component failure, and are also the "trace" left by the defect, which is equivalent to a "fingerprint"; the failure mechanism is the physical, chemical and other damage evolution processes experienced inside the material when the component fails, which is the key to failure analysis; the failure cause is the key factor leading to component failure, that is, the root cause, which is the purpose of failure analysis. We usually use in-depth fracture analysis, including macro analysis, meso analysis and micro analysis, to accurately find the source of defects, determine the damage evolution process and its mechanism, and trace the cause of failure, such as fatigue, overload, impact, manufacturing defects, improper workmanship, poor design, etc.
The failure causes are also divided into root causes, secondary causes and related causes, and ABCs are listed in order of severity. What really needs to be clarified is the root cause. If only the superficial cause is found and the root cause is not traced, similar failure accidents will still happen again in the future.
In addition, for a complex failure case, failure analysis generally involves nine steps: on-site investigation, appearance inspection, performance test, microscopic analysis, environmental media analysis, simulation analysis (including bench test), comprehensive analysis, analysis conclusion and solution.
Later
: After finding the cause, how can we confirm that the modification plan is really effective?
Professor Yang Zhenguo of Fudan University
: After the improvement plan is implemented, physical tests must be conducted under actual working conditions to quickly identify whether the improvement plan is effective. As far as material performance analysis is concerned, testing is generally divided into three levels. The first level is self-inspection, where you verify your own modification plan; the second level is mutual inspection, where other personnel or similar institutions conduct inspections to see if the results are consistent and compliant; the third level is external inspection, which is inspection by an independent third-party inspection agency. Other inspections are authoritative. Self-inspection alone does not constitute an integrity assessment, and different levels must be tested and identified.Later
: Can modifying the national standard and specifying a higher pedal load capacity solve the problem?
Supplier controller development engineer: This number is difficult to directly define as a unified national standard value because the braking structure is always changing. If the future is full wire braking, no longer relying on mechanical connections, and full braking can be achieved by applying a force of several hundred N through the simulator, then there may be no requirement for the original force value. Many of the figures in the industry are formed through long-term practice.
Brake engineer of independent car company
: Many of these indicators come from the car companies’ own experience. Whether it costs 3000 N or how much, each company may be different. You cannot break away from the specific structure and just compare a number.Researcher on Materials Mechanics at Tsinghua University
: Currently, there is still a lack of sufficiently suitable and comprehensive evaluation standards for non-metal pedal brackets. According to the original method of evaluating metal stents, it is impossible to evaluate the characteristics of non-metallic materials so comprehensively. Material selection changes, and testing and inspection standards must be adjusted accordingly.Purchasers from joint venture car companies
: I don’t care about the specific numerical value. This is a matter of research and development. I care about the impact of this matter. Risk analysis has three dimensions: severity, frequency of occurrence, and discoverability. This severity is close to infinity. If the national standard cannot be covered, then the national standard must be changed and the tester will be called over to calibrate the new standard on site.Later
: As an ordinary consumer, what can you do?
Foreign Tier 1 brake system R&D engineer
: Ordinary consumers basically can’t tell the difference. We who do research and development may have some industry experience considerations when buying a car, but consumers without relevant experience have a hard time judging how well the design and verification are done from the appearance, configuration, or a test drive.Cars at different price points will indeed differ in terms of supplier selection and specific design, but the price may also include a high brand premium, so being expensive does not necessarily mean that there will be no problems. For individuals, failure rate is a big data concept. But once I encounter it, it's 100%. A consumer may only buy a car every five years and only two or three in a lifetime. Big data is of very limited help to individual consumers.
Later
: Consumers can’t verify it themselves. What information should companies at least disclose?
Researcher on Materials Mechanics at Tsinghua University
: It should explain what materials are used in the parts, whether the test method is specific to the characteristics of this material, and how to supplement the verification for parts not covered by the original standard. Materials and construction have changed and it's not enough to say "meets old standards".Source of title picture: "Speed of Life and Death"
Comments