♦Two radiators can share one OEM number and still differ in life and performance. A look at aluminum choice, tube and fin tooling, brazing and leak testing at TONGSHI.
Look at a radiator and the geometry seems almost too simple: flat tubes and fins brazed into a core, a pair of tanks, a handful of brackets and necks. That simplicity is misleading. Once a unit is in service — heat cycling, vibration, coolant chemistry, stone impact, pressure spikes — the material choices and the process discipline behind that simple shape decide whether it lasts a season or a decade.
A simple shape, a much less simple build
The core is the working part. Coolant flows through flat tubes, fins are folded between them, and both are joined to header plates that distribute flow into the tanks on either side. Everything else — side plates, inlet and outlet necks, mounting brackets, drain cock — either supports that core or connects it to the vehicle.
Each of those parts can be made well or made cheaply, and the difference rarely shows at the counter. It shows up as a pinhole at a tube-to-header joint after two winters, a fin pack that has collapsed and closed the airflow, or a side plate that has bowed under thermal cycling and started loading the core it was meant to protect. This is why material selection and process control are treated here as two halves of one argument rather than two separate chapters.
Material first: why aluminum, and which aluminum
Radiator cores have historically come from two material families: copper-brass and aluminum. Copper is the better conductor, and a brazed or soldered copper-brass core tolerates a neglected cooling system well and can usually be repaired — which is why it still serves certain heavy-duty and older applications. The penalty is mass and cost. A copper-brass core of the same capacity is considerably heavier than an aluminum one, and the soldered assembly carries more labour and more joints. For modern passenger cars and light commercial vehicles, that combination no longer makes commercial sense.
Aluminum is roughly one third the density of copper, so the weight saving is immediate and it compounds through the whole vehicle: less load on the mounting points, less vibration energy fed into the core, easier service access. Its conductivity is lower than copper in absolute terms, but a heat exchanger is designed around surface area, wall thickness and air-side geometry rather than around bulk conductivity alone. Aluminum can be formed into thinner tube walls and denser fin packs, and the clad alloys used for brazing give it corrosion protection as part of the wall rather than as a coating on top of it. The finished core closes most of the conductivity gap and does it at lower mass and lower cost. It also suits the manufacturing route: welded and brazed in a single controlled-atmosphere furnace cycle, and fully recyclable at end of life.
Where quality is actually decided: four process steps
Material sets the ceiling. Process decides how much of that ceiling the finished part actually reaches.
Tubes: layered alloy, welded seam
The tube starts as a 4343/3003/7072 clad aluminum strip with a single brazing layer on the outside, formed and welded into a flat tube by high-frequency welding. Each layer has a job. The 4343 Al-Si layer is the brazing clad that will melt and form the joint. The 3003 Al-Mn core carries the strength and the pressure rating. The 7072 Al-Zn layer on the coolant side acts as a sacrificial anode and protects the tube from the inside out.
Fins: folded edges and a pitch held at 2.9-3.0 mm
Fins are produced on a set of folded-edge forming rolls. The folded edge is what gives the fin its resistance to collapse: when the core is clamped, brazed and then pressurized in service, a plain louver fin buckles easily and closes the air path, while a folded edge holds its shape and keeps the fin pack open.
Header plates: plugged corners, ribbed teeth
The header plate is where the core meets the tanks and where pressure loads concentrate. The tube ends at all four corners of the plate are plugged, which relieves the outer tubes on both sides of the core and raises the pressure the assembly can carry.
Side plates: relief grooves against thermal growth
Side plates are made from the same 4343/3003/7072 specification, in inserted and outer-hung versions. Both carry relief grooves along their length. A radiator grows and shrinks with every heat cycle; with nowhere for that movement to go, the side plate bows, the core distorts and the joints fatigue. The grooves absorb the movement, which is why a part that looks like a bracket is one of the reasons a core still holds its shape after years of cycling.
Quality control: three gates that prove the build
Brazing: why a core is furnace-brazed and not simply assembled
A radiator only works if heat can pass from the tube into the fin. Two metal surfaces pressed together are not enough: the interface between them acts as a thermal resistance, and under the temperatures and vibration a cooling system sees, that contact degrades. Brazing removes the interface. The assembled core passes through a controlled-atmosphere furnace under nitrogen, the 4343 clad melts at roughly 577-600 °C while the 3003 core is still solid, and capillary action draws the filler into every fin-to-tube and tube-to-header joint in a single cycle.
Three things come out of that one pass: a metallurgical bond that carries heat, a rigid core that resists pressure and vibration, and leak-tight joints. The process is fluxless, so no corrosive flux residue is left behind inside the coolant path.
Red-water test: making a small leak visible
Pressure testing with dyed water is the second gate. The core is filled and held under pressure with water carrying a red tracer, then inspected. Plain water can pass a leak of a few microns without showing anything; the dye leaves a visible red trace at the point of seepage, so a joint that only leaks when pressurized can be found and removed before it reaches a box.
Gas detection: finding what water cannot
Helium detection is the third gate. Helium is inert, it does not react with coolant, and its molecule is small enough to pass through a leak that would never show a water bead; a mass spectrometer picks up the trace and locates it. Nitrogen covers the pressure-hold and decay checks — dry, inert, and it leaves nothing behind inside the core.
Between them, the red-water test proves the unit holds pressure at a level the eye can verify, and the gas test proves it holds pressure at a level water cannot resolve. A core that passes both is packed dry, clean and verified, rather than probably fine.
What this means when you specify or buy
An OEM number is a fitment promise. It tells you the part will bolt in and the hoses will line up. It says nothing about the alloy, the clad layer, the fin pitch, the braze yield or the leak test behind it. That is why two suppliers quoting the same number can be quoting two different products at two different prices — and why the cheaper one is not always the cheaper one across a service life.
If you buy radiators, the questions worth asking are concrete: what alloy and clad specification are the tube and the fin, how is fin pitch controlled, what brazing method and atmosphere are used, what leak test is applied and at what pressure, what braze pass rate is sustained, and how is the side plate relieved. A supplier who can answer those questions has a process. One who cannot has a price.
TONGSHI has manufactured cooling products since 2004 from its base in Fogang, Qingyuan, Guangdong: radiators, condensers, intercoolers, heaters, oil coolers and evaporators across 58 series and more than 18,000 part numbers for passenger cars, commercial vehicles and construction machinery. The material and process decisions described above are the same ones behind every one of those numbers.
Send an OEM number, a drawing or a sample, and we will tell you what we would build, how we would test it, and what it would cost.
