Everything between the turbocharger compressor outlet and the engine inlet is a pressurised path, made of rigid tubes, flexible couplings, clamps and at least one heat exchanger. Its job is to move compressed air without losing it. Every joint in that path is a potential leak, and there are usually more joints than a customer counts when they picture the system.
The path normally includes a compressor outlet connection, a hot-side pipe, an intercooler inlet tank, the core, the outlet tank, a cold-side pipe, a throttle body connection, and often a blow-off or bypass valve with its own flange. Each of those interfaces has a seal, and each seal has to hold under conditions that change constantly while the car is being driven.
The simplest leak test in a workshop is static. The system is capped, air or smoke is introduced through a port, and the operator looks for escaping gas or listens for it. It is a fast and useful check, and it catches the gross faults: a hose that is not seated, a clamp that was forgotten, a blanking plug missing from a spare port.
What a static check does not do is reproduce working load. On the road or on a dyno, the system experiences pressure that rises and falls with turbo output, temperature that climbs as the air is compressed, and mechanical movement as the engine rocks on its mounts. A joint can seal perfectly at rest and leak as soon as those three arrive together, which is why a static test that comes back clean is not the end of the investigation.
Materials move at different rates when they get hot. An aluminium pipe grows more than the silicone coupling around it, and a stainless flange grows less than the tank it is welded to. A clamp that holds a cold joint perfectly can be at a different tension once the joint reaches working temperature, and a hose that grips its spigot firmly when cold can slide as the metal under it expands.
There is also a stiffness change. Silicone softens as it warms, so a coupling that resists pressure well when cold can begin to deflect outward as it heats, allowing the liner to lift away from the bead. And the pressure itself is not steady: closing the throttle against a turbo that is still spinning sends a sharp transient through the whole path, which is the moment an already marginal joint lets go.
Experience puts leaks in a fairly consistent set of places. The clamp line at the end of a hose is the most common, particularly where the clamp sits slightly off the bead or where two clamps are too close together. Sensor bosses and spare ports on a fabricated tank or pipe come next, followed by the weld seams where a tank meets an end plate, and then the joints at a valve flange.
Intercooler cores leak in a different way. An impact or a fatigue crack in the core itself produces a leak that shows up as a loss of pressure without an obvious external escape path, and it is often found only after the rest of the system has been ruled out. Because the core and its end tanks are made and welded as one unit on our intercooler line, that joint is where our own inspection concentrates before a unit ships.
The couplers themselves deserve particular suspicion. A soft hose with the wrong bore, a bead that is too shallow for the clamp being used, or a clamp tightened until the liner is bruised will all produce intermittent leaks that look like something else entirely. Our performance silicone hose range is built around that interface rather than around the tube alone.
Most leaks are found with simple methods, applied in a sensible order. A soap solution brushed along each joint while the system is held under a modest static pressure will bubble at the leak point, and it localises faults well enough to move on. Smoke introduced into the path is useful for the same reason and is much easier to see in a crowded engine bay.
Listening is underrated. A leak that only appears under load usually makes a sound: a hiss, a whistle or a fluttering noise that changes with throttle position. A driver's description of where and when a noise occurs is often more valuable than a workshop inspection, because it tells the technician which condition to reproduce.
Visual inspection after a hard run also helps. Oil mist carried through the intake path leaves a witness mark at a leaking joint, and a clamp that has been moving will show a bright polished ring on the spigot. Those traces survive after the leak has stopped, which makes them a good starting point on a system that is not currently misbehaving.
Testing a charge system while it is actually working means holding the car in a condition where the turbo is producing output and the load is being absorbed, then observing whether the system holds. That requires either a rolling road or a road test with an observer, and it requires instrumentation: at minimum a way to see whether the pressure the engine builds is the pressure a healthy system of that kind should build.
A drop in measured pressure for a given engine demand, compared with the same car in a known good state, is the classic signature of a leak. The difficulty is that intake restriction, compressor condition, ambient temperature and intercooler heat soak all move the same measurement. Isolating a leak from those effects is what turns a reading into a diagnosis, and it is usually done by comparing the same car against itself at a different time rather than against another car.
A leak-down test under load tells you whether a system held at the moment it was tested, under the conditions of that test. It does not tell you that the system is sealed for all conditions. A marginal joint can hold throughout a moderate session and fail on a hot day, on a long climb, or on the first cold morning after a repair.
That limit argues for a different way of reading the result. Instead of treating a clean test as a pass, treat it as one data point, and combine it with the physical inspection of the joints, with the condition of the clamps and beads, and with the driver's account of the behaviour. The combination is far more reliable than any single reading, and it is how a workshop narrows down an intermittent fault.
Most charge-side leaks in custom work can be designed out at the specification stage. The following points are the ones that determine whether a system is easy to seal.
Getting these right reduces the amount of chasing required once the car is assembled. Rigid charge pipes are covered under titanium welding parts, intake-side hardware under air intake pages, and tanks and brackets under other parts.
After the chasing is done, the answer is often less exotic than expected. A clamp that was tightened onto the bead rather than behind it. A coupling reused one job too long, with a liner that has taken a set. A pipe whose two ends were sited so that engine movement pulls one joint apart while the other is pushed together. None of those are material problems, and none of them are fixed by changing material.
The productive habit is to treat the charge path as an assembly whose joints are as important as its pipes, and to test it in a condition as close to real use as the workshop can manage. Everything else, including the choice of what the pipes are made from, is decided through the same made-to-order conversation described on our customization pages.
