Air Bearings and Precision Air Support

Air Bearings and Precision Air Support

Required asset: Air Bearing Supply Acceptance Brief

R01 decision: Verify cleanliness, pressure stability and documented equipment limits.

R01 release gate: Use bearing-maker limits; no universal air-bearing quality class.

1 Direct answer

An air-bearing or precision-air support must be supplied to the exact bearing maker’s documented inlet-pressure, flow, cleanliness, temperature, pressure-ripple and start-up/shutdown requirements. Acceptance should be demonstrated at the bearing supply interface during the governing motion, load and competing-demand events. Compressor-room pressure or a generic air-quality label does not prove the installed bearing receives suitable air.

It does not create a universal cleanliness class, pressure margin, restrictor size, lift gap, load capacity or shutdown rule.

2 Controlled scope

It covers aerostatic or other externally supplied air-bearing arrangements only to the extent supported by exact equipment documentation.

Bearing selection, pad geometry, stiffness, damping, load capacity, motion accuracy, structural dynamics, control-loop design and machine safety remain with the bearing maker and responsible machine disciplines.

4 Why point-of-use evidence matters

An air bearing operates through a thin pressurised gas film and restricted flow paths. Small changes in inlet pressure, contamination, temperature or restriction can affect lift, stiffness, flow and position. The applicable sensitivity must come from the exact bearing documentation.

Pressure can be stable at the main header while dipping after a filter, regulator, small tube or local valve. Air can meet a treatment claim at the plant but acquire water, particles, oil or debris through downstream piping and maintenance.

5 System boundary

Draw the boundary from the nominated plant-air tie-in through isolation, dryer, filters, receiver if present, distribution, final filter, regulator, valve, tubing, manifold and bearing inlet. Identify exhaust or venting, purge branches, monitoring points and temporary connections.

Define whether one measurement represents one pad, a manifold, an axis or the full machine. Record other users sharing the branch and any stored volume inside the accepted boundary.

6 Operating states

7 Pressure definitions

Record gauge, absolute and differential pressure. Bearing-maker inlet limits normally apply at a named interface; do not substitute regulator setpoint or upstream header pressure. If a manifold feeds several elements, identify the pressure point represented.

Static pressure before and after a move can miss transient ripple or dips. Use a measurement system with range, response and sample rate compatible with the maker’s acceptance requirement.

8 Reference flow

Record whether flow is reference volume, standard volume, free-air delivery, actual compressed volume or mass flow. Preserve reference pressure, temperature, humidity and standard. A bare L/min or CFM value is ambiguous.

For a measured reference trace q_ref(t), event quantity is the integral over time. Mean flow can support utility accounting, while instantaneous or band-limited behaviour may govern bearing stability.

9 Bearing-maker limits

Create a requirement ledger that quotes or paraphrases the exact maker document, model, revision and interface. Include minimum and maximum inlet pressure, flow basis, required cleanliness, allowable temperature, ripple or stability criterion, sequencing, shutdown and alarm requirements where supplied.

If a field is absent, record Unknown and obtain clarification. Do not reverse-engineer an approval limit from a catalogue graph or a different bearing family.

10 Bearing-specific cleanliness

Cleanliness can include particle size/count, water content or pressure dew point, oil aerosol, liquid oil, vapour or other contaminants. Requirements depend on restrictor geometry, materials, process sensitivity and environment.

A generic class is not automatically applicable. Record every dimension of the maker’s requirement and the sampling location. Treatment equipment ratings must be reconciled with installed inlet conditions and maintenance state.

11 Particle control

Particles can obstruct small restrictions, damage surfaces or change flow distribution. Final filtration location, rating definition, integrity, bypass risk, housing cleanliness and element condition matter.

Record commissioning flush or cleaning requirements from authorised sources. Filter installation itself can introduce debris; downstream work needs contamination controls.

12 Water and dew point

Liquid water, condensation and ice can obstruct paths or alter performance. Pressure dew point must be stated with pressure and measurement basis. Ambient or surface temperature determines whether condensation risk exists downstream.

Record worst credible local temperature, shutdown cooling and depressurisation effects. Dryer outlet evidence does not prove the bearing inlet if distribution conditions differ.

13 Oil and vapour

Oil requirements must follow the bearing maker and process. Compressor technology alone does not establish delivered oil condition. Lubricated-system carryover, hydrocarbon vapour, downstream materials and maintenance products may be relevant.

14 Final filtration and pressure loss

Final filters can protect the bearing but add pressure drop that increases with flow, contamination and element condition. Record clean and service-limit pressure loss where applicable, differential-pressure monitoring and replacement criteria.

Assess dynamic pressure downstream of the final treatment during governing demand. Raising plant pressure is not a substitute for correcting a restricted filter.

15 Regulation

A regulator can show acceptable no-flow pressure yet droop or oscillate at changing flow. Record inlet, outlet, command or setpoint, downstream volume, bleed, range and response evidence.

Select regulation from maker requirements and installed dynamics.

16 Pressure ripple

Pressure ripple is variation about the required supply condition. Define metric, frequency range, observation window and location from the bearing-maker or machine requirement. Peak-to-peak, RMS and minimum pressure are not interchangeable.

Correlate ripple with compressor controls, regulators, valves, other users, machine motion and measurement-system response. Do not smooth away relevant components.

17 Thermal stability

Gas temperature, machine structure, bearing surfaces and ambient conditions can affect dimensional stability. Compression, pressure drop, purge, warm-up and changing room conditions can alter local temperature.

Record inlet-air temperature, ambient, bearing or structure temperature where required, warm-up rule and stability window.

18 Tubing, manifold and local volume

Tube bore, length, fittings, manifolds and flexible lines affect pressure loss and response. Unequal branches can produce unequal pad conditions. Local volume can buffer or delay changes.

Record installed routes and modifications. A nominal connector size does not establish the effective flow path.

19 Competing demand

A shared branch can experience actuator strokes, blow-off, tool use, other bearings or maintenance connections. Define the plausible coincident event and measure it rather than adding every connected nameplate flow.

The required worked test observes pressure at the bearing during a competing event. It must preserve machine state, load, motion, branch configuration and timestamps.

20 Start-up support

The maker or machine design may require air before load, release, motion or control enable. Record prerequisite pressure, stability, purge, temperature and confirmation signals. A pressure switch setpoint alone may not prove the full readiness condition.

Validate sequence and timing under the authorised commissioning method.

21 Shutdown support

Controlled shutdown may require motion to stop, load to transfer, mechanical support to engage or surfaces to separate before air decays. Name the required sequence, minimum retained condition and confirmation evidence.

Loss-of-air response is distinct from normal shutdown. Both require approved safe-state definitions.

22 Monitoring and alarms

Identify pressure, differential pressure, flow, temperature, dew point or contamination monitoring required by the maker or project. Record sensor range, uncertainty, sample rate, filtering, alarm delay and action.

An alarm does not replace a protective design. Avoid universal thresholds.

23 Measurement methodology

Define the question, exact bearing/interface, machine load, motion, state, competing event, instruments, timebase, sample rate, validity rules and uncertainty. Measure upstream and at the bearing where necessary to separate plant variation from local loss.

Preserve raw data, configuration, calibration, sensor ports, tubing, timestamps and processing. Temporary instruments must not contaminate or materially restrict the supply.

24 Time alignment

Use a shared clock or documented offset for bearing pressure, upstream pressure, competing-user command, machine motion, load and alarms. Account for controller and acquisition latency.

A causal claim needs time resolution better than the transient. Preserve unsmoothed data.

25 Pressure budget

A pressure budget names the required bearing inlet condition and each measured or calculated loss through treatment, distribution, regulation, valves and final connections under the governing flow.

26 Flow and inventory calculations

For constant reference flow q_ref during t seconds, V_ref = q_ref × t / 60. For a fixed local volume V changing between absolute pressures p1 and p2, a simplified inventory screen is Delta V_ref = V × (p2 − p1) / p_ref under stated temperature assumptions.

These equations do not size the bearing, predict film behaviour or approve local storage.

27 Concurrency and multiple bearings

For multiple axes or machines, build a chronological event register. Sum only actual coincident flows. Distinguish continuous bearing flow from transient purge, lift or auxiliary demand.

Average demand supports plant accounting; minimum pressure and ripple during coincidence govern installed acceptance where the maker requires them.

28 Worked Case A — pressure at the bearing during competing demand

Assumptions. Fictional demonstration only. The bearing maker’s real limit is unknown. During precision motion without a competing event, upstream pressure is 6.4 bar gauge, bearing inlet pressure is 5.8 bar gauge and measured ripple is 0.06 bar peak-to-peak. A separate actuator then operates for 1.2 seconds on the shared branch. Upstream pressure falls to 6.0 bar gauge, bearing inlet falls to 5.1 bar gauge and ripple reaches 0.18 bar peak-to-peak. Recovery to within 0.05 bar of the 5.8 bar baseline takes 1.7 seconds.

Method. Align upstream pressure, bearing inlet pressure, competing-event command and motion state. Compare endpoint loss and ripple before, during and after the event.

Result. Baseline upstream-to-bearing loss is 0.6 bar. Event minimum loss is 0.9 bar. Upstream droop is 0.4 bar; bearing droop is 0.7 bar; the additional local loss is 0.3 bar. Ripple increases by 0.12 bar peak-to-peak.

Interpretation. The trace supports both an upstream supply change and increased local loss. It does not establish acceptance because maker limits are absent. Investigate shared supply and local treatment/regulation separately.

29 Worked Case B — continuous and transient volume

Assumptions. A fictional bearing consumes 24 L_ref/min continuously for a 10-minute observation. A start purge uses 60 L_ref/min for 12 seconds.

Result. Continuous volume is 240 L_ref. Purge volume is 12 L_ref. Total is 252 L_ref.

Interpretation. The purge controls a short peak but contributes less volume than continuous support. Do not add purge flow continuously.

30 Worked Case C — multiple-bearing concurrency

Assumptions. Four fictional bearing branches each use 18 L_ref/min continuously. One purge uses 50 L_ref/min and another machine event uses 90 L_ref/min. The two transient events can align or be sequenced.

Result. Continuous demand is 72 L_ref/min. Aligned peak is 212 L_ref/min. With non-overlapping permitted events, peak is 162 L_ref/min. Total event volumes remain unchanged.

Interpretation. Scheduling can change the shared peak but cannot override start-up or safety sequencing.

31 Worked Case D — pressure budget

Assumptions. Fictional upstream pressure is 6.5 bar gauge. Measured losses during governing flow are 0.15 bar through central treatment, 0.20 bar through distribution, 0.35 bar through final filtration/regulation and 0.10 bar through final tubing.

Result. Calculated bearing inlet is 5.70 bar gauge and total loss is 0.80 bar.

Interpretation. The calculation reconciles measured losses but does not state acceptability. Compare with the exact maker limit.

32 Diagnostics

33 Commissioning

Confirm model, revision, maker requirements, supply boundary, treatment, filters, regulator, tubing, manifold, sensors, alarms, safe states and machine sequence. Verify cleanliness evidence before connecting or exposing sensitive paths.

Under the authorised method, record pre-start, pressurisation, lift/support, idle, governing motion, competing demand, hold and shutdown. Preserve deviations and retest evidence.

34 Acceptance

Acceptance must name bearing model, maker-document revision, load and motion case, inlet-pressure limits, flow basis, cleanliness dimensions, temperature range, ripple metric, monitoring, competing event, start-up/shutdown sequence and approving owners.

The release gate requires maker-specific limits. Unknown maker limits force HOLD; a generic air-quality class cannot close the item.

35 Frequently asked questions

Can I specify one ISO air class for every air bearing?

No. Use the exact bearing-maker requirements and every relevant contamination dimension.

Does a clean compressor guarantee clean bearing air?

No. Treatment, piping, maintenance and final connections determine delivered condition.

Can stable header pressure prove acceptance?

No. Measure at the bearing interface during governing events.

Will a receiver solve ripple?

Not automatically. Source controls, regulator dynamics, restriction, local volume and event timing all matter.

Can the worked pressure be used as a setting?

No. All values are fictional and the maker limit is deliberately unknown.

What happens when limits are missing?

Record Unknown and HOLD; obtain the maker’s documented requirement.

36 Required asset — Air Bearing Supply Acceptance Brief

Complete with actual project evidence. Use Unknown for missing facts. Dispositions: DRAFT, READY FOR REVIEW, HOLD and ACCEPTED FOR STATED DUTY.

A Identity and maker requirements

B Supply and treatment

C Operating cases and measurements

D Instruments and evidence

E Acceptance disposition

F Filled fictional illustration

37 Claim and uncertainty limits

Worked values are fictional. They establish no bearing rating, load capacity, film thickness, precision, cleanliness class, pressure setting, flow, temperature band, margin, customer result or certification.

Measurement uncertainty can include calibration, response, sensor port, sampling, reference conversion, timing, repeatability and processing. Use the method required by the maker or project.