11.1 Cavitation, Flashing, Choked Flow, and Noise
Key Takeaways
- Flashing occurs when downstream pressure P2 is below liquid vapor pressure Pv, so vapor that forms at the vena contracta persists instead of collapsing.
- Cavitation occurs when vena-contracta pressure drops below Pv but P2 recovers above Pv; bubble collapse pits metal downstream of the restriction, not upstream of it.
- High-recovery valves such as ball and butterfly styles have a lower liquid pressure-recovery factor FL and cavitate at a smaller ΔP than low-recovery globes at the same P1, P2, and Pv.
- Hardened trim delays erosion; multi-stage or low-recovery trim is what keeps local pressure above Pv. Flashing is a lasting phase change and is not cured by anti-cavitation cages.
- Convert P1 and P2 to absolute pressure before comparing them with Pv. Comparing a psig downstream reading with a psia vapor pressure is a standard exam error.
11.1 Cavitation, Flashing, Choked Flow, and Noise
free PE Control Systems practice questionsPractice questions with detailed explanations
A control valve is a variable restriction. Flow accelerates, static pressure falls, and a vena contracta forms just downstream of the throttle: the smallest effective flow area, the highest velocity, and the lowest static pressure Pvc. Everything in this section is a comparison of Pvc and recovered P2 with Pv. Use absolute pressures for that comparison.
Pressure profile: flashing versus cavitation
Walk the static-pressure path:
- Upstream of the valve: P1.
- At the vena contracta: Pvc, the minimum.
- Downstream, after some recovery: P2.
Flashing is a phase change that persists. If P2 < Pv, vapor that forms at the vena contracta does not collapse back to liquid. The outlet is a high-velocity two-phase stream. Metal loss looks like wire-drawing, grooving, or smoothing on the outlet and downstream pipe. That is erosion by a lasting mixture, not the pockmarked pitting of bubble collapse.
Cavitation is formation and collapse. If Pvc < Pv but P2 > Pv, bubbles form and then implode when pressure recovers. Collapse occurs downstream of the vena contracta — typically on the downstream face of the plug and seat, the cage windows, and the body wall just after the restriction. Microjets from collapsing cavities pit the surface. Operators often hear gravel in the pipe.
No vaporization requires Pvc > Pv (and therefore P2 > Pv). A modest ΔP on a low-recovery globe can stay in this region. The same ΔP on a high-recovery butterfly may not.
Do not treat “any vapor in the valve” as cavitation. If the vapor stays vapor at P2, you are in flashing. If it collapses, you are in cavitation. Damage location follows that difference.
Choked flow is a capacity ceiling
Choked flow is not a damage name. It is the point where further lowering P2 does not increase mass flow.
For liquid, once the vena contracta is essentially at vapor pressure, the valve is choked. Additional ΔP is wasted as noise, vibration, and (if P2 is still above Pv) cavitation collapse. For gas or vapor, choke is sonic velocity at the vena contracta when the P2/P1 ratio falls below the fluid’s critical pressure ratio (near 0.5 for diatomic gases; use the fluid’s heat-capacity ratio k when a problem gives it).
On the exam, choked liquid flow is the reason a huge extra ΔP does not buy you more gallons per minute. Size from the choked ΔP, not from the full P1 − P2 you see on the P&ID.
Worked decision: given P1, P2, and Pv
Convert gauge readings to absolute before comparing with Pv.
Case A. P1 = 120 psig, P2 = 20 psig, Pv = 40 psia.
- P1 = 120 + 14.7 = 134.7 psia
- P2 = 20 + 14.7 = 34.7 psia
- P2 (34.7 psia) < Pv (40 psia) → flashing.
Do not specify anti-cavitation multi-stage trim as if it will “unflash” the outlet. You need velocity control, hard facing, an angle body that sends the jet into the pipe, sacrificial downstream spool, or a process change that raises P2 above Pv.
Case B. P1 = 120 psig, P2 = 60 psig, Pv = 40 psia.
- P2 = 74.7 psia, which is above 40 psia → not flashing.
- Cavitation is now the question: will Pvc drop below 40 psia? If ΔP is large and the valve is high-recovery, treat it as cavitation until a manufacturer σ or FL check says the vena contracta stays wet.
Case C. Same P1 and P2 as Case B, but a cold liquid with Pv = 5 psia. Flashing is off the table. Cavitation margin is much larger, but a high-recovery rotary can still dip Pvc below 5 psia on a very large ΔP.
Exam trap: comparing P2 in psig with Pv in psia, or calling Case A cavitation because “bubbles form in the valve.”
σ and FL — qualitative, not a formula sheet
The liquid pressure recovery factor FL (ISA/IEC control-valve sizing practice) measures how close Pvc is to P2:
- Low-recovery valves (globe, angle, multi-stage cages): FL is typically high, on the order of 0.85–0.95. Pressure does not rebound much, so Pvc is not far below P2. Harder to cavitate for the same P1, P2, and Pv.
- High-recovery valves (ball, butterfly, many eccentric plugs): FL is typically lower, often about 0.55–0.75. Pressure crashes at the vena contracta and then recovers toward P2, so Pvc can be far below P2. Same ΔP, more cavitation.
A cavitation index σ is a dimensionless margin, commonly built from (P1 − Pv) relative to ΔP. Large σ means vapor pressure is far below upstream pressure compared with the drop — more margin. When σ approaches a manufacturer inception value, noise starts, then damage. You do not need a memorized numerical σ limit; you do need to know that high recovery (low FL) and low σ travel together.
Choked liquid ΔP is on the order of FL²(P1 − Ff), where Ff brings Pv (and critical pressure) into the ceiling. Qualitatively: smaller FL → choke and cavitation at a smaller ΔP.
Joule–Thomson, two-phase feed, and noise
Joule–Thomson throttling is isenthalpic expansion through the restriction. For most gases near ambient temperature the Joule–Thomson coefficient is positive, so temperature falls. Hydrogen, helium, and neon can warm on expansion in common plant conditions. Exam uses: hydrate risk on natural-gas lets, ice on a regulator, CO2 cold spots, and “the gas cooled even though we did no shaft work.”
Two-phase feed or a flashing outlet means density is no longer the liquid ρ you used in a single-phase Cv calculation. Mass flow drops relative to an all-liquid prediction, noise rises, and erosion follows the high-velocity mixture. Do not size a flashing hydrocarbon as if it stayed liquid all the way to P2.
Noise has two families:
- Hydrodynamic: liquid turbulence and especially cavitation collapse (gravel sound).
- Aerodynamic: compressible mixing and shock cells downstream of a gas restriction.
Path treatment (heavier wall, acoustic insulation, silencers) treats the observer. Staging, low-noise cages, and lower velocity treat the source. A silencer on a flashing outlet does not stop metal loss.
Mitigation that matches the physics
Hardened trim (stellite facing, 440C, tungsten carbide) buys life. It does not raise Pvc. Use it for residual cavitation or flashing erosion after you have reduced severity.
Staging (multi-step trim) splits ΔP so no single stage drives Pvc below Pv (or below a damage σ). That is the real cavitation fix.
Prefer a low-recovery globe or angle when ΔP is high relative to (P1 − Pv). A butterfly on a pump minimum-flow bypass is a classic cavitation victim.
| Phenomenon | What you observe | Where metal is lost | Mitigation that actually matches |
|---|---|---|---|
| Cavitation | Gravel noise, vibration, pitted “cinder” surface | Downstream of the vena contracta (plug/seat/cage/body after the throttle) | Low-recovery or multi-stage trim; reduce ΔP; hardened trim only as residual protection |
| Flashing | High outlet velocity, two-phase mist, may be quieter than cavitation | Outlet and downstream pipe (erosion, wire-drawing) | Raise P2 if possible; angle body; hard facing; sacrificial spool; do not expect anti-cavitation cages to collapse vapor that cannot collapse |
| Choked liquid | Flow no longer rises as P2 falls | Same as cavitation if P2 > Pv; flashing pattern if P2 < Pv | Size on choked ΔP; change FL or stages; do not “buy Cv” with extra ΔP |
| Choked gas | Sonic limit; further ΔP does not raise mass flow | Aerodynamic noise, vibration, possible shock damage | Low-noise trim, staging, heavier downstream pipe |
| Joule–Thomson cooling | Outlet temperature drop on gas | Hydrates, ice, brittle seals — not pitting | Heat tracing, material/elastomer rating, hydrate inhibition |
| Two-phase | Capacity shortfall versus liquid Cv; erosion | Trim and downstream | Two-phase sizing method; velocity and hardness — not a liquid-only Cv |
A liquid hydrocarbon control valve has P1 = 160 psia, P2 = 45 psia, and Pv = 70 psia. Which phenomenon should you expect?
Where does damaging cavitation typically attack a control valve?
For the same P1, P2, and Pv, which valve is more likely to cavitate?