Metallic ringing in the minutes after the engine stops is thermal contraction, and it is normal. The exhaust cools from the manifold end first and the rear last, so different sections contract at different rates. Where a thin shield or a pipe is restrained, that contraction is released in small steps, and each step is heard as a tick.
The point is that this is a shutdown event, not a running fault. If the noise only happens after the engine stops and nothing is heard while it runs, the system is behaving as designed. A fault present while running is a different diagnosis and should not be confused with this one because both involve hot metal.
Pitch tells you about the part. A large thin panel rings low and long; a small bracket ticks sharply and briefly. Heat shields are the usual suspects because they are thin, clamped at a few points and close to the hottest parts. One that has lost a fastener rings more, being free to vibrate.
Wait until the car is cold, then tap each shield with a fingertip. A well-fixed shield gives a dull, damped sound; one that rings freely is loose. Do not tap a hot shield, and do not use a metal tool, because a thin shield dents easily and a dent creates a new contact point.
Normal contraction noise is brief, happens once and fades as the system cools. A fault produces a ring that repeats or continues long after the metal should be cool, or one accompanied by a smell, a glow or a change in how the engine runs. Persistence is the useful signal here.
Location matters too. Contraction noise comes from wherever the system is restrained, so it tends to come from the same place each time. A ring that moves around, or that comes from a part with no reason to be hot, is not contraction. An exhaust heatblanket changes how a section cools.
Insulation changes the cooling rate, and cooling rate changes the sound. A wrapped section holds heat longer, so it contracts later and more slowly, which can spread the ticking out or silence it. A shield that traps air against a pipe can also create a small resonant cavity that rings higher than the pipe alone.
That is why fitting a shield or blanket sometimes introduces a noise rather than removing one. The part is doing its job, but the air gap is a new acoustic element. Clearance between shield and pipe is the variable to control: too close and it buzzes, too far and it stops shielding.
There is one case worth attention. A crack at a weld rings differently from a healthy joint, because the two faces are free to vibrate against each other as they cool. The sound is sharper and more irregular than the even tick of a contracting shield, and it may carry a faint rasp.
Look for a discoloured line at the weld toe and check it again after a few heat cycles. A crack grows, so a mark that gets longer is a crack and not a stain. This is the one post-shutdown noise worth investigating early, because a crack at a hanger or flange weld is easier to deal with while it is short.
In most cases, nothing. If the noise is brief, comes from a shield and stops when the system is cool, it is doing what hot metal does. Check that all shield fasteners are present and that the shield is not touching the pipe, then leave it. Wrapping a section usually just changes the pitch.
Where the noise is genuinely intrusive, improve the mounting rather than adding insulation. A system restrained at too many points ticks at each one; one with a single fixed point and free movement elsewhere ticks less. That is why a stainless steel exhaust system built with deliberate hanger placement is quieter at shutdown.
