How a turbocharger works.

A spiral casting — a volute — whose cross-section shrinks as it wraps around the wheel. That taper is the whole design: gas is bled off to the blades continuously as it travels around, and the falling area keeps the remaining gas moving at the same speed instead of stalling. It runs at exhaust temperature, around 900 °C at full load, which is why it is cast iron rather than aluminium.

• The A/R ratio — scroll area divided by the radius it acts at — is the number that sets the trade between spool and top-end flow
• A small A/R accelerates the gas harder, so boost arrives early but the housing chokes sooner
• Twin-scroll housings split the flow into two passages so the exhaust pulses from different cylinders cannot interfere with each other

In the workshop: Cracks in the scroll between the wastegate port and the flange are normal on a high-mileage unit and are usually harmless. Cracks that reach the wheel bore are not.

A radial-inflow wheel: gas enters at the rim, turns through ninety degrees and leaves along the axis, giving up pressure and temperature as shaft work on the way. It is cast in a nickel superalloy — Inconel or similar — because it sits in gas that can exceed 950 °C while turning fast enough that the blade tips are close to the material's limit on their own. It is welded to the shaft and the pair are balanced together as one assembly.

• More blades than the compressor wheel, because the gas arrives with far more energy and needs turning more sharply
• Its mass is most of the rotating inertia, which is most of the reason turbo lag exists
• Tip clearance is a few tenths of a millimetre — a rub here is what a badly worn bearing eventually causes

In the workshop: Nicked or missing blade tips mean something went through the engine. Replacing the turbo without finding out what will simply destroy the next one.

The axial exit from the turbine, and the flange the downpipe bolts to. Gas leaving here has given up most of the pressure it arrived with, but it is still hot, and how freely it can escape matters more than most people expect: back-pressure at the outlet subtracts directly from the pressure ratio across the turbine, and therefore from the drive available to the compressor.

• A restrictive downpipe or a blocked catalyst raises turbine outlet pressure and costs boost everywhere
• This is the one place on a turbo system where a larger pipe reliably helps
• Outlet temperature is always lower than inlet temperature — the difference is the work the turbine took out

In the workshop: If a car makes less boost than it should and the exhaust smells hot at the manifold, suspect the downpipe or the catalyst before touching the wastegate.

A pressed steel disc behind the turbine wheel, clamped between the turbine housing and the bearing housing. It keeps the radiated heat of the exhaust off the bearing housing and its oil, and it keeps heat in the exhaust, which is not incidental: hotter gas carries more energy to the turbine. Outside the turbo, blankets and pressed shields do the same job for the actuator diaphragm, the oil lines and anything in the engine bay that does not enjoy being next to a component that glows.

• Stands between gas at around 900 °C and oil that must stay well below its coking temperature
• Retaining heat in the gas raises turbine inlet temperature and helps spool
• In the teardown it comes off after the shaft and turbine wheel slide out

In the workshop: Outside the turbo, leaving the engine-bay shields off is how actuator diaphragms and oil feed lines fail early.

The entire machine is two turbomachines on one shaft, and this is the shaft. Turbine power flows in at one end and compressor power flows out at the other; the difference, minus bearing drag, is what accelerates or decelerates the rotating assembly. The simulator solves exactly that equation thousands of times a second, and at steady state the turbine's power equals the compressor's plus the bearings' to within 1%, which the test suite checks.

• Turns at up to about 180,000 rpm on the mid frame, or 3,000 revolutions a second
• Welded to the turbine wheel; the compressor wheel is clamped on with a nut
• Balanced as a complete assembly — a wheel cannot be swapped on its own

In the workshop: Shaft play is checked in two directions and they mean different things: a little rocking is normal, but movement along the axis means the thrust bearing has gone.

The centre housing rotating assembly — the part sold as a cartridge when a turbo is rebuilt. It holds the bearings, the oil galleries that feed them and the seals at each end, and it has to do that with red-hot iron bolted to one face and cold aluminium to the other. Most centre housings are also water-cooled, which is less about running temperature than about what happens after shutdown, when there is no oil flow and the heat in the turbine housing has nowhere to go but inward.

• Carries a continuous flow of engine oil that both lubricates and cools
• Water jackets on most modern units prevent oil coking after a hot shutdown
• The temperature gradient across it is one of the harshest in the whole engine

In the workshop: The reason for an idle-down after a hard drive is this part. Stop the engine hot and the oil sitting in these galleries bakes into carbon, which then blocks them.

Most turbochargers do not use ball bearings. They use fully floating plain bushes that spin freely in the housing, so there are two oil films — one between shaft and bush, one between bush and housing — and the bush turns at roughly half shaft speed. Nothing touches anything: the shaft is held on a pressurised wedge of oil, which is why oil pressure is not a maintenance concern but a structural one.

• Two oil films instead of one halves the shear rate and roughly halves the drag
• A ball-bearing cartridge spools faster and costs more, which is why it is a performance option
• Lose oil pressure for a second at full speed and the bearings are gone

In the workshop: Almost every turbo failure that is not an object through the wheel starts at these. In the cutaway they turn at half the shaft's speed, as real ones do.

The turbine is pushed one way by exhaust pressure and the compressor the other way by boost, and the two do not balance. What is left is a steady axial load, and this flat bearing takes it. It is the part that suffers when a compressor works hard against a restriction — a blocked air filter or, worse, a throttle that has just slammed shut on full boost, which is the load case the recirculation valve exists to prevent.

• Wears fastest under high boost with low flow — exactly the surge condition
• End play is the standard diagnostic: measured in hundredths of a millimetre
• Failure here lets the wheels touch their housings, which ends the turbo

In the workshop: If a turbo has audible flutter on every gearchange, this is the part quietly paying for it.

At each end of the centre housing sits a ring in a groove, working like a miniature piston ring. They are not tight seals and are not meant to be — they rely on gas pressure being higher outside than inside to keep oil where it belongs. That is why turbo oil leaks so often appear at idle or on the overrun, when there is little boost on the compressor side to hold the oil back, and why a blocked oil drain or a pressurised crankcase causes smoke that looks like a failed turbo but is not.

• They work by pressure balance, not by sealing tightly
• Blue smoke on overrun is the classic symptom of the compressor-side ring losing that balance
• A restricted oil drain floods the housing and pushes oil past them even when they are perfect

In the workshop: Not drawn separately. Before condemning a turbo for smoking, check the crankcase breather and the oil drain — both cause identical symptoms and neither needs a new turbo.

Not drawn separately on this model.

A small line from the engine's oil gallery to the top of the centre housing. It carries the film the bearings float on and a large share of the cooling: oil leaves a turbocharger considerably hotter than it arrives. Because it feeds a plain bearing at extreme speed, what matters is continuous pressure — a momentary interruption at 150,000 rpm does damage that no amount of subsequent oil will undo.

• Fed from the same gallery as the engine bearings, so low oil pressure hurts both
• Usually the smallest-diameter oil line on the engine, and the easiest to block with debris
• Restrictors are sometimes fitted on ball-bearing units, which need less flow

In the workshop: After any rebuild, prime the turbo with oil before cranking — the first few seconds are otherwise a dry start at speed.

A large, near-vertical pipe from the bottom of the centre housing back to the sump. It has no pump behind it: oil leaves by gravity, and it leaves as a hot froth of oil and air that needs far more room than its volume suggests. This is why the drain is several times the diameter of the feed, why it must not run uphill anywhere along its length, and why a turbo mounted below the oil level in the sump cannot be drained at all without a scavenge pump.

• Must be larger than the feed and must fall continuously to the sump
• A restricted or uphill drain floods the housing and forces oil past the seals
• Crankcase pressure works directly against it, so a blocked breather has the same effect

In the workshop: When a freshly fitted turbo smokes, the drain routing is the first thing to check and the most common thing to have got wrong.

A billet or cast aluminium wheel that does the machine's actual job. Air enters along the axis, is accelerated outward by the blades and leaves the rim fast; the housing around it then converts that speed into pressure. Compressing air heats it — unavoidably, and by more the harder the wheel works, which is what the intercooler downstream exists to undo. Notice on the model that every second blade is shorter: those are splitters, added so the inlet is not choked by blade metal where the passage is narrowest.

• Aluminium, because at this end the gas is cold and the priority is low inertia
• The inducer diameter — the bore the air enters through — is what sets the wheel's flow capacity
• Its efficiency is the readout that decides how much of the compression became pressure rather than heat

In the workshop: A wheel with all its leading edges eroded has been eating dirt. That means the air filter or the pipework upstream of it, not the turbo.

The intake bore ahead of the wheel, shaped to let air turn into the inducer without separating from the wall. Everything upstream of this point is under vacuum, not pressure, which changes what goes wrong: joints do not blow off, they suck in unmetered air and dirt. It is also where the recirculation valve returns its charge, because putting that air back in front of the wheel is what keeps flow through the wheel during a lift-off.

• The only part of the induction system that runs below atmospheric pressure
• A collapsed or restrictive intake pipe costs boost the wastegate cannot get back
• The recirculation valve plumbs back in here rather than venting to atmosphere

In the workshop: Dirt on the inducer blades with a clean filter means the leak is between them. Check the filter housing seal and every clamp before blaming the filter.

The cold-side volute — a mirror image of the turbine housing, with its section growing rather than shrinking as it wraps around. Air leaves the wheel rim fast and at modest pressure; the widening passage slows it down, and a slowing gas in a closed duct raises its pressure. This is where a large part of the pressure ratio on the compressor map is actually produced, and it is why housing choice changes a compressor's behaviour as much as wheel choice does.

• The diffuser — the gap between the wheel rim and the start of the scroll — does the first stage of that conversion
• Cast aluminium, because the air here is warm rather than hot
• Some housings carry a ported shroud, a slot that widens the surge margin at low flow

In the workshop: The bright ring you sometimes find rubbed inside a compressor housing is the wheel having touched it. That is a bearing verdict, not a housing one.

Left alone a turbocharger is positive feedback: more boost burns more fuel, which makes hotter exhaust, which spins the turbine harder, which makes more boost. Something has to interrupt that loop before it destroys the engine, and this is it — a flap over a port in the turbine housing that lets exhaust gas past the wheel without doing any work on it. It opens as soon as boost reaches its target and throws away exactly as much drive as it takes to hold the number. Set it to Held shut in the simulator, or run the No wastegate experiment, and you can watch the loop do what it does without one.

• The flap opens further as engine speed rises — holding the same boost costs more bypass at higher flow
• It regulates by wasting energy, which is why a wastegated turbo is never at peak efficiency on boost
• An external wastegate is the same idea in a separate casting, used where the internal port cannot flow enough

In the workshop: A turbo that overboosts at high rpm usually has a wastegate that cannot flow enough, not one that is stuck. Watch the flap: if it is already fully open, the gate has run out of authority.

A sealed can holding a spring and a rubber diaphragm, with boost pressure piped to one side. When pressure times diaphragm area beats the spring, the rod moves, turns the lever, and the flap cracks open — a purely mechanical proportional controller, and the reason the simulator's Auto mode opens the gate in proportion to the boost over its target rather than anything cleverer. The spring sets the lowest boost the system can hold; an electronic control valve bleeding pressure away from the can is how a modern engine raises the target above it.

• Spring pressure sets the minimum boost — you cannot tune below it without changing the spring
• A boost control solenoid works by lying to the actuator about how much pressure there is
• The diaphragm is the least heat-tolerant part on the hot side, which is what the heat shield protects

In the workshop: Check it with a hand pump, not by guesswork: the rod should start to move at the spring pressure and be at full travel a few tenths of a bar above it.

A small piston valve on the compressor housing, also called a diverter or bypass valve. When the throttle slams shut on boost, the air in the charge pipe has nowhere to go but the compressor is still spinning, so the flow through the wheel stalls and reverses: surge. The valve reads the pressure drop across the throttle, lifts its piston, and lets the trapped charge flow back to the compressor inlet, so air keeps moving through the wheel. A blow-off valve does the same job but vents to the atmosphere, which is where the famous sound comes from.

• Opens on the pressure difference across the throttle, not on boost alone
• Returning the air to the inlet keeps it metered, which matters on engines that measure air flow before the turbo
• Take it off and a lift-off surges: the Lift off experiment runs both

In the workshop: A turbo that flutters on every lift-off either has a torn valve diaphragm or none at all. The thrust bearing is what pays for it.

The plumbing between the compressor outlet and the inlet manifold, by way of the intercooler. Its volume matters more than it looks: everything downstream of the compressor is a plenum that has to be filled before boost appears at the engine, so a long charge path adds directly to the lag. The simulator models this explicitly as a filling equation — the compressor delivers, the engine takes, and the pressure between them is what reconciles the two.

• A bigger intercooler and longer pipework buy cooler air at the cost of slower response
• Every joint here is under pressure — a boost leak vents straight to atmosphere
• This volume is what stays pressurised when the throttle shuts, and therefore what surges

In the workshop: Boost leaks are quiet and cost more than people expect. Pressurise the whole charge path cold with a leak tester and listen; do not go looking on a running engine.

Not drawn separately on this model.

Compressing a gas heats it, and hot air is less dense — so a compressor that made a bar of boost has also thrown away part of what that boost was for. The intercooler is a heat exchanger that gives the heat back before the air reaches the cylinders. It is not an optional efficiency accessory: cooler charge is denser charge, and it also raises the knock limit, which is what lets a boosted engine run enough ignition advance to use the air it has been given.

• Denser charge means more oxygen for the same manifold pressure
• Cooler air resists knock, which is often worth more than the density gain
• It costs a little pressure to push air through, so a bigger core is not automatically better

In the workshop: Judge one by outlet temperature and pressure drop together. A core that cools brilliantly but eats a fifth of a bar getting there has given back what it saved.

Not drawn separately on this model.