On October 8, a Chinese automotive media outlet published a video of an extreme braking test: three Zunjie V800 MPVs, tested from 100 km/h to a full stop, and in all three the brake pedal bracket fractured at the moment the pedal hit the floor. The brakes reportedly did not survive four hard applications.
That evening the manufacturer responded that no such failure had occurred in real customer use since first deliveries, and offered a free upgrade for the part. The stock reaction was blunter than the public argument: JAC Motor hit its daily limit down, then again, leaving a market value around RMB 50.1 billion.
There are two layers to the dispute. On the surface it is about test methodology — how far continuous extreme braking is from everyday driving. Underneath it is a structural question: the brake pedal is the last mechanical link between the driver and the braking system. Under what conditions does its load-bearing part fail, is there any warning before it does, and how much margin was left?
We are not doing a technical appraisal and we are not taking sides. What interests us, as a company that has spent ten years landing product structure and process into production, is the question that keeps getting discussed and keeps getting executed badly: when you shave structure and material to make a product lighter and cheaper, how far can you go before you cross the safety line?
Weight reduction moves stress around
Injection-molded parts typically lose weight along three routes. All three remove mass. All three redistribute stress inside the material, and each has a different safety cost.
Route 1: thinner walls. The simplest way to raise stiffness is to add thickness; the most direct way to lose weight is to remove it — and this is where the risk concentrates. Load capacity is not linear in section size. For a beam in bending, stiffness scales roughly with the cube of thickness. Take a wall from 3.0 mm to 2.4 mm and you drop about 20% of the weight while stiffness can fall by more than half.
Molding adds its own constraints. Thinner walls mean higher melt flow resistance, harder filling, and a markedly higher chance of short shots, incomplete filling and weak weld lines. To fill a thin wall you usually raise injection pressure and melt temperature, which pushes up residual stress and warpage. Thinning a wall is never a single number change; it is a chain of process parameters being re-sequenced.
Route 2: glass-fiber reinforcement. Swapping ordinary engineering plastic for PA66 with 30% glass fiber is the most popular "thinner without losing strength" answer. The fiber does raise tensile strength and modulus, which in principle permits thinner walls. But there is a trap that does not show up in the datasheet: glass-fiber reinforced plastic is anisotropic. Fibers orient strongly along the flow direction, and strength along the flow can be more than double the strength across it. A rib whose direction is perpendicular to fiber orientation may deliver only half the strength the datasheet implies. Many failures happen near weld lines, because the two sides of a weld line have fibers oriented perpendicular to each other, making the interface the weakest location in the part.
Glass-fiber parts are also notch-sensitive: under impact or stress concentration they crack more readily from sharp corners, screw bosses and sharp edges than unreinforced material. So you buy higher modulus with lower toughness and stricter fillet requirements.
Route 3: ribs, cutouts and topology optimization. This is the most orthodox approach — remove material where stress is low, add ribs along the load path, build a load-bearing skeleton from the least material. The direction is right, but its validity depends heavily on two things: a real load spectrum, and how well the simulation matches the physical part. If the load case is estimated from normal driving while the real extreme condition is several times that, the optimized skeleton has been carved against a wrong premise — the safety margin never existed, it just had not been triggered yet. Ribs have side effects too: an abrupt wall-thickness change at the rib root creates a stress concentration, and ribs that are too dense or too thick cause sink marks and localized residual stress during cooling, which can become crack initiation sites.
Four things worth reviewing
1. Load-bearing safety parts need their own stop line
In vehicle and appliance development, parts are split into load-bearing safety parts and general structural parts. When a general structural part fails at weight reduction, the result is deformation, noise or shortened life. When a safety part crosses the line, the result is irreversible. A brake pedal bracket is a textbook load-bearing safety part.
Its development logic cannot be driven by "how many grams did we save". It has to run the other way: first define the minimum safe section from the extreme load case, then optimize within that section — not set a weight target and try to buy strength back afterwards.
Typical safety factors: metal load-bearing structures usually 1.5 to 2.5, and plastic parts higher because of creep and aging. The critical point is that the factor applies to the worst-case load, not the average. A safety factor built on average pedal force protects the average — not the extreme.
2. Plastics lose strength over time — a bill metal parts do not have
If the load-bearing part is engineering plastic, you owe two extra calculations. Creep: under sustained constant load, plastic deforms slowly and continuously, so a structure that passed at the factory can develop loose assemblies and redistributed stress years later. Fatigue: under repeated cyclic load, plastic life curves are steeper than metal and sensitive to temperature and humidity.
Which means validation of a plastic load-bearing part cannot stop at "did it pass one extreme test". It has to answer "how much margin is left after five years". Substituting a one-off test on a brand-new part for a whole-life safety judgment is the most common methodological gap in disputes like this.
3. Weld lines and fiber orientation are invisible strength black holes
A drawing can specify wall thickness, rib positions and fillets precisely, but it will not say "the weld line lands exactly where stress is highest". A weld line is the interface where two melt fronts meet and fuse; the lower the fusion temperature and the shorter the time, the weaker the interface — typically 60% to 80% of bulk strength, with a bigger penalty in glass-fiber material.
This is why gate location is structurally part of the design, not something to adjust after the mold is cut. The right practice is to run mold-flow analysis during design and overlay weld lines, air traps and fiber orientation onto the stress contour, confirming the weld line is not in a high-stress region. On weight-reduction projects this step is especially non-optional, because after thinning the whole stress level rises and positions that used to be safe can become critical.
4. "Extreme test" versus "real-world use" is really about missing validation layers
Both sides of this argument are right, because the answer depends on which question you are asking. Everyday use cares about "it will not break under normal pedal force". Safety design cares about "it will not fail in an extreme case". Those are two different validation layers.
A mature validation system usually has three: durability under normal conditions, load capacity under extreme conditions, and failure mode and effects analysis. The key move in the third layer is to actively imagine "what happens if this part breaks" and check whether there is redundancy or any warning mechanism. For a braking system, failure of a mechanical connection means the direct interruption of pedal force transmission — the highest severity class — and it should be classified as a safety-critical characteristic with tightened controls.
What designers can take away
For the three weight-reduction routes, the safety cost differs: wall thinning affects stiffness most; fiber reinforcement brings anisotropy and notch sensitivity; structural optimization is relatively controllable but still needs re-verification of margin under extreme load.
The key is separating load-bearing safety parts from general structural parts. A safety part's weight reduction must not back-derive strength from a weight target — define the minimum safe section from extreme load first, then optimize within it. General structural parts can be weight-target driven. Applying one weight-reduction metric to every part is a common source of this class of risk.
FAQ
Does injection-molding weight reduction weaken product safety?
It depends which route and how far. Of the three, wall thinning hurts stiffness most, fiber reinforcement introduces anisotropy and notch sensitivity, and structural optimization is relatively controllable — but all of them require re-verifying safety margin under extreme load.
Why can't I just read glass-fiber plastic strength off the datasheet?
Because the strength is anisotropic: along the fiber orientation it is markedly higher than across it, sometimes by a factor of two. The datasheet gives an ideal-orientation value, while in a real part orientation is set by flow. On top of that, fiber orientations on either side of a weld line are perpendicular, so the interface is the weakest point in the part — usually 60% to 80% of bulk strength.
What safety factor do load-bearing safety parts use?
Metal load-bearing structures commonly 1.5 to 2.5, plastic parts higher because of creep and aging. The important part is that the factor is computed against the worst-case load, not the average.
Is mold-flow analysis necessary on weight-reduction projects?
Yes, especially after thinning. Thinning raises the overall stress level, so a weld line position that was safe can become critical. Mold-flow analysis gives you weld line, air trap and fiber orientation distributions ahead of time, and the correct practice is to overlay them on the stress contour and confirm the weld line avoids high-stress regions. Done at the design stage, it costs far less than reworking after the mold is cut.
Top comments (0)