A diaphragm pressure gauge measures fluid pressure by sensing the elastic deflection of a thin metal disc clamped at its edge, with a linkage converting that deflection into pointer movement on a calibrated dial. Typical accuracy runs ±1.6% to ±0.5% of full scale per ASME B40.100, with ranges from 0–10 mbar up to 0–40 bar. The diaphragm form solves what a Bourdon tube cannot — low-pressure, viscous, slurry, or sanitary service. You see them on dust collector baghouses, dairy CIP loops, and chemical metering skids.
Pressure Gauge (form 2) Interactive Calculator
Vary diaphragm bow, full-scale calibration, motion ratio, and dial sweep to see the amplified linkage travel and pointer rotation.
Equation Used
The diaphragm center bow, delta, is multiplied by the movement ratio R to estimate linkage travel. The dial pointer angle is calibrated linearly from the measured bow relative to the full-scale bow, delta_FS, over the selected full-scale sweep theta_FS.
- Diaphragm deflection is linear within the elastic measuring range.
- The linkage and gear train follow the stated motion ratio with negligible backlash.
- The over-range stop is represented at 110% of full-scale travel.
Operating Principle of the Pressure Gauge (form 2)
A diaphragm gauge works on a simple idea — a flat or corrugated metal disc, clamped between two flanges, bows outward when pressure pushes on one side. That bow is tiny, often less than 0.5 mm at full scale, but a stiff link pinned to the diaphragm centre drives a sector gear and pinion, which multiplies the motion into pointer rotation across a 270° arc. The whole movement sits in a sealed case, often filled with glycerine or silicone oil to damp pointer flutter on pulsating service.
Why use this form instead of a Bourdon tube? Two reasons. First, a Bourdon tube needs significant pressure to flex meaningfully — below 0.6 bar it gets sloppy and hysteretic. A diaphragm reads cleanly down to a few millibars. Second, the wetted side can be a flat sanitary face, flush with the process, with no internal pockets where product sits and rots. That matters for dairy, pharma, and food. Run a Bourdon tube on yogurt and you're growing a science experiment inside the C-tube within a week.
Tolerances matter. The diaphragm thickness sets the spring rate, and a 10% thickness error shifts full-scale deflection by roughly 30% because stiffness scales with thickness cubed. If you over-pressurise a low-range gauge — say a 0–60 mbar unit hit with 2 bar — the diaphragm yields plastically and the zero shifts permanently. You'll see the pointer rest below zero on the next cycle and never come back. Other common failure modes are diaphragm fatigue cracking after 10⁶+ pressure cycles, link-pin wear that shows up as 2–3% reading drift, and oil-fill leakage that lets the pointer flutter on pump pulsation.
Key Components
- Sensing diaphragm: A thin elastic disc, typically 50–100 mm diameter and 0.05–0.3 mm thick, made from 316L stainless, Hastelloy C-276, or tantalum for aggressive media. Corrugations are coined into the disc to linearise deflection across the range. Thickness tolerance must hold ±5% or full-scale accuracy drifts outside the ±1.6% class.
- Upper and lower flanges: Clamp the diaphragm rim with controlled bolt torque — typically 15–25 Nm on M8 fasteners — to seal the process side without distorting the active area. The lower flange carries the process connection, usually 1/4 NPT, 1/2 NPT, or a tri-clamp sanitary fitting.
- Push rod and movement: A stiff stainless link transfers diaphragm centre travel to a sector gear pivoting on jewel or bushing bearings. Gear-to-pinion ratio is usually 60:1 to 120:1, multiplying 0.4 mm of diaphragm travel into ~270° of pointer sweep.
- Hairspring: Removes backlash from the gear train by pre-loading the pinion in one direction. Without it, the pointer would lag 1–2% on falling pressure compared to rising — a textbook hysteresis signature.
- Liquid fill (optional): Glycerine 99.7% or silicone oil fills the case on pulsation-damping versions. The fill also keeps moisture out at sub-zero ambient. A glycerine fill turns cloudy and yellow at 5+ years — that's your visual cue the gauge has aged.
- Over-range stop: A mechanical stop limits diaphragm travel to ~110% of full scale to prevent plastic yield on accidental over-pressure. Quality gauges hold 5× over-range without zero shift; cheap ones yield at 1.3×.
Industries That Rely on the Pressure Gauge (form 2)
Diaphragm gauges show up wherever Bourdon tubes fail — low pressure, sticky media, sanitary service, or aggressive chemistry. The flush wetted face and clean elastic deflection sensor combination makes them the default low-pressure measurement choice across process instrumentation. You'll find them in pulp mills, breweries, baghouse monitoring, and chemical dosing skids where the pressure transducer alternative is overkill or too fragile.
- Sanitary food & dairy: Anderson-Negele SA sanitary diaphragm gauges on Tetra Pak CIP return lines, 3-A certified with 1.5" tri-clamp connection and flush 316L face.
- Dust collection / HVAC: Dwyer Series 2000 Magnehelic differential gauges across pulse-jet baghouses on Donaldson Torit dust collectors, reading 0–4 inH₂O across the filter bank to trigger pulse cleaning.
- Chemical dosing: WIKA 432.50 chemical seal gauges on Grundfos DDA digital dosing pumps metering sodium hypochlorite into a municipal water plant — the PTFE-coated diaphragm survives chlorine where a brass Bourdon would pit through in months.
- Pulp & paper: Ashcroft 1005 phosphor-bronze diaphragm gauges on Voith stock prep dilution lines, where 4% consistency pulp slurry would plug a Bourdon tube siphon within hours.
- Pharmaceutical bioreactors: Rosemount 1199 diaphragm seals paired with mechanical gauges on Sartorius BIOSTAT D-DCU fermenter headspace, reading 0–500 mbar gauge with autoclavable flush diaphragm.
- Natural gas distribution: Winters PFP Series gas service diaphragm gauges on regulator station inlets, reading 0–60 psi with internal blowout disc to UL/cUL gas service spec.
The Formula Behind the Pressure Gauge (form 2)
Diaphragm gauge designers and instrument users both need to predict centre deflection as a function of applied pressure, because that single number sets the full-scale travel that drives pointer sweep. At the low end of the range — say 10% of full scale — deflection is tiny and any link-pin slop dominates the error budget. At nominal full scale you're at peak linearity, the sweet spot the gauge was calibrated for. Push beyond 110% and the diaphragm leaves the elastic regime, the formula stops applying, and you eat permanent zero shift. The classical equation below is the small-deflection plate result for a clamped circular diaphragm.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| yc | Centre deflection of the diaphragm | m | in |
| P | Applied pressure differential across the diaphragm | Pa | psi |
| R | Active radius of the diaphragm (inside the clamped rim) | m | in |
| t | Diaphragm thickness | m | in |
| E | Young's modulus of the diaphragm material | Pa | psi |
| ν | Poisson's ratio of the diaphragm material | dimensionless | dimensionless |
Worked Example: Pressure Gauge (form 2) in a semiconductor wet-bench HF acid line
A wafer fab in Hsinchu is fitting a PTFE-faced diaphragm gauge to a hydrofluoric acid recirculation loop on a Modutek wet bench. The line normally runs at 350 mbar (0–700 mbar gauge full scale), but they need to know predicted diaphragm centre travel so they can spec the correct linkage and confirm the gauge will read cleanly at the 70 mbar low-flow alarm point and survive the 700 mbar full-scale condition. Diaphragm: 316L stainless, R = 25 mm, t = 0.15 mm, E = 193 GPa, ν = 0.30.
Given
- R = 0.025 m
- t = 0.00015 m
- E = 193 × 10⁹ Pa
- ν = 0.30 dimensionless
- Pnom = 35,000 Pa (350 mbar)
- Plow = 7,000 Pa (70 mbar)
- Phigh = 70,000 Pa (700 mbar)
Solution
Step 1 — collapse the constant terms first so we only sweep pressure across three operating points:
Step 2 — nominal operating point at 350 mbar (35,000 Pa):
That nominal deflection is healthy — well above the 0.4 mm threshold where link-pin slop starts dominating, comfortably below the corrugated-diaphragm linear limit of about 1/3 the diaphragm thickness for a flat plate or several mm for a properly corrugated disc. The pointer will track smoothly across mid-scale.
Step 3 — low-end alarm point at 70 mbar (7,000 Pa):
At 0.72 mm of travel you're at 10% of full scale — readable, but resolution is tight. A 50 µm link-pin clearance shows up as a 7% reading wobble at this point. That's why low-flow alarms on critical service get a separate transmitter rather than relying on the gauge alone.
Step 4 — full-scale at 700 mbar (70,000 Pa):
7.16 mm flat-plate prediction exceeds the small-deflection assumption (which holds to roughly t = 0.15 mm of travel for an uncorrugated disc). In practice you spec a corrugated diaphragm here, which linearises behaviour out to several mm, and you add a mechanical over-range stop at ~7.5 mm to protect against the 1.5 bar hammer transients you'll see when the recirc valve slams shut.
Result
Predicted nominal centre deflection at 350 mbar is 3. 58 mm — a comfortable mid-scale travel that gives a clean, readable pointer position around the 12 o'clock face. The full sweep runs from 0.72 mm at the 70 mbar low-flow alarm up to 7.16 mm at 700 mbar full scale, which tells you immediately the diaphragm must be corrugated rather than flat to stay linear at the top end. If the gauge reads 20% low compared to a reference transducer, the most likely causes are: (1) diaphragm thickness above spec — a 0.17 mm disc instead of 0.15 mm cuts deflection by roughly 30% because stiffness scales with t3, (2) push-rod buckling from a previous over-pressure event, visible as a permanent negative zero offset on vent, or (3) wetted-face fouling by HF reaction product building a stiff fluoride layer that locally raises diaphragm spring rate.
Pressure Gauge (form 2) vs Alternatives
The diaphragm form competes mainly with Bourdon-tube gauges for general service and with electronic pressure transducers for tight-accuracy or remote-readout work. Each wins on different axes — pick by service conditions, not by habit.
| Property | Diaphragm gauge | Bourdon tube gauge | Electronic pressure transducer |
|---|---|---|---|
| Usable low-pressure range | 0–10 mbar to 0–40 bar | 0.6 bar full scale minimum (sloppy below) | 0–1 mbar to 0–1000 bar |
| Accuracy class (typical) | ±1.6% to ±0.5% full scale | ±1.0% to ±0.1% full scale | ±0.1% to ±0.025% full scale |
| Wetted-face cleanability | Flush, sanitary, no dead pockets | Internal C-tube, traps product | Depends on diaphragm seal option |
| Survives slurry / viscous media | Yes, primary use case | No — siphon plugs in hours | Only with diaphragm seal accessory |
| Over-range tolerance | 3–5× full scale typical | 1.3× full scale before yield | 1.5–2× full scale before sensor damage |
| Cost (industrial 100 mm dial) | $80–$400 | $25–$150 | $200–$1,500 plus power and readout |
| Cycle life before drift | 10⁶–10⁷ cycles | 10⁶ cycles typical | 10⁸+ cycles, no mechanical wear |
| Power required | None — purely mechanical | None — purely mechanical | 12–24 VDC loop power |
Frequently Asked Questions About Pressure Gauge (form 2)
Almost always over-torqued clamp. A tri-clamp sanitary fitting only needs hand-tight plus a quarter turn — maybe 15 Nm equivalent. Crank it down with a wrench and you compress the gasket so hard it bows the lower flange and pre-stresses the diaphragm inward. The diaphragm now sees its zero point at a partially deflected state and reads low across the whole range.
Diagnostic check: loosen the clamp to finger-tight, vent the line to atmosphere, and watch the pointer. If zero jumps up by 2–5%, you found it. Re-torque to spec with a calibrated tri-clamp wrench, not a pipe wrench.
Direct-wetted is fine when the gauge body material itself can take the chemistry — 316L for most aqueous service, Hastelloy C-276 for chlorides, tantalum for hot HCl. You go to a chemical seal (a separate isolating diaphragm with capillary-coupled fill fluid) when (a) you need a non-standard wetted material like PTFE-lined or gold-plated that's only available as a seal, (b) the process temperature exceeds the gauge movement rating of about 60°C, or (c) you want field-serviceable gauge replacement without breaking process containment.
The trade is response time and accuracy. A chemical seal adds 0.5–2 seconds of lag and typically 1% extra error from fill-fluid thermal expansion. On a slow process loop nobody cares. On a fast control loop, that lag will drive a PID controller into oscillation.
Two usual suspects. First, the case is dry-fill but the service has high-frequency pulsation — reciprocating pump, solenoid valve cycling, or compressor ripple. The diaphragm and movement have no damping and resonate. Swap to a glycerine-filled version or add a snubber.
Second, a glycerine-filled gauge that lost its fill through a cracked window or leaking fill plug now behaves like a dry gauge but with a partially flooded movement, which is worse than either. Look at the fill level — there should be a small bubble, about 10% of case volume, sitting at the top. No bubble at all means thermal expansion will pop the window. No fill at all means refill or replace.
Rule of thumb from ASME B40.100: pick a full-scale range so normal operating pressure sits between 25% and 75% of full scale, ideally around 50–60%. Below 25% you're working in the noisy bottom end where link-pin slop dominates and the pointer barely moves. Above 75% you have no headroom for upset conditions and you'll yield the diaphragm on the first water-hammer event.
For a 350 mbar normal operating pressure, that means a 500 mbar or 700 mbar full-scale gauge — not a 400 mbar gauge that will sit at 88% and not a 1.6 bar gauge that will sit at 22%.
Diaphragm gauges accumulate fatigue damage faster than Bourdon tubes when run near full scale. Each pressure cycle works the elastic disc through a strain range, and after 10⁵ to 10⁶ cycles you start seeing measurable zero shift before any visible cracking. If your service cycles every few minutes — like a batch fill or a CIP rinse — you'll rack up a million cycles in a year or two.
The fix is either to oversize the gauge so cycles happen at 30–40% of full scale instead of 80%, or to spec a higher-fatigue-life diaphragm material like Inconel 718 or Hastelloy. Don't expect a $90 commodity gauge to outlive a Bourdon on heavy cycling service — it physically can't.
The textbook small-deflection formula assumes a flat, uncorrugated disc and a deflection less than about one-third of the disc thickness. Real industrial diaphragms are corrugated — coined with concentric ridges — to extend the linear range and increase stiffness in a controlled way. Corrugations add bending stiffness through a geometric factor that the flat-plate formula does not account for, typically 1.3 to 2× depending on corrugation depth and count.
For design work you either use the manufacturer's published P-vs-y curve, run a finite-element model with the actual corrugation geometry, or apply an empirical corrugation factor. The flat-plate formula is the right starting point for sanity-checking a thin smooth diaphragm, not for predicting absolute deflection on a coined production part.
References & Further Reading
- Wikipedia contributors. Pressure measurement. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
