Motion Platform Maintenance: Keeping 6-DOF Bases Honest

A motion platform leaves the factory meeting its specification — repeatability of +/-0.02 mm on the YT-6DOF, response under 10 ms on the simulator bases. Whether it still meets that specification in year three is a maintenance question. This post sets out the upkeep rhythm for the platform classes in the motion platform line.

What wears, by drive type

Servo electric cylinders — the drive on most simulator bases here — wear at ball screws, bearings and encoder couplings. The published advantage is cleanliness (no oil contamination risk) and precision (+/-0.1 mm positioning on the racing platform), and the maintenance profile follows: lubrication schedules, backlash checks, and encoder health monitoring rather than fluid management.

Hydraulic drives — used where loads and accelerations demand them (the tank trainer's +/-2G base, 400 kg class tables) — add fluid, filters, seals and hose inspection to the calendar, plus oil analysis for high-duty installations. Hydraulic accuracy is excellent but it is rented from the maintenance programme.

Electromagnetic drives — the high-frequency micro-vibration option — shift wear to electronics and cooling.

Calibration is a scheduled event

Closed-loop feedback (encoders, IMU on the rotational platform) keeps motion honest day to day, but the loop itself needs verification: pose accuracy checks against a reference, repeatability runs, and delay measurements. Treat the published figures — +/-0.005 mm repeatability on the 3-axis positioning platform, +/-0.1 micrometre translation on the rotational platform — as acceptance values to re-verify annually, with the method agreed at purchase.

Design features that pay for themselves

• Modularity — the 3 DOF platform publishes quick replacement of sensors and actuators; a failed actuator becomes a swap, not a rebuild.
• Fewer joints — three-axis platforms carry half the actuator count of six-axis; the maintenance budget follows the joint count.
• Standard interfaces — EtherCAT/Modbus diagnostics beat proprietary black boxes when a fault must be found at 7 a.m. before a training day.
• Published spares posture — the manufacturer states spare-parts service for delivered systems; confirm the wear-parts list (seals, screws, encoders) and lead times at contract, not at first failure.

A sane service calendar

Daily: visual and noise checks by operators, fault-log review. Monthly: lubrication points, fastener torque sampling, hydraulic fluid level and filter state where fitted. Quarterly: full-stroke motion tests, backlash and delay spot checks, safety circuit tests. Annually: calibration verification against published accuracy, actuator wear assessment, control-system backup and update. Written this way, maintenance is a page in the operations manual — and the platform you bought is the platform you still have in year five.

Fault diagnosis: the order that finds causes

When a platform misbehaves — a knock, a lag, a pose error — diagnose in the order the energy flows. First, the command path: is the control system asking for the motion correctly (interface logs, EtherCAT/Modbus diagnostics)? Second, the feedback path: do encoders and IMU agree with the commanded pose, or is the loop chasing a lying sensor? Third, the drive path: actuator current draw and temperature against baseline. Fourth, the mechanical path: backlash, lubrication, fastener torque. Most 'platform faults' are feedback faults wearing mechanical clothes, and the order above finds them before a spanner is lifted.

Software and control upkeep

Modern platforms are software with steel attached. Keep control firmware and motion libraries under version control, back up configurations before every update, and never update control software the week before a major training or operating commitment. The custom interfaces on the Pro-class platforms (customised industrial protocols) make documentation part of the asset: the integrator's interface document should live with the platform's maintenance file, and its loss should be treated as a defect.

Records, training and the maintenance file

Every platform should carry a living maintenance file: commissioning acceptance values (the published accuracy figures as measured at handover), the service calendar's completion record, fault history with root causes, calibration certificates, and the spares inventory. Train at least two staff members on the platform's daily and monthly checks — a maintenance regime that lives in one person's memory has a single point of failure exactly where the safety case cannot afford one. When the file is complete, an auditor, an insurer or a buyer of the business can see the platform's life at a glance; that visibility is itself worth money.

Spares strategy and obsolescence

Platform spares divide into three classes. Consumables — lubricants, filters, seals on hydraulic units — are bought on schedule and held locally without debate. Wear parts — ball screws, bearings, encoder assemblies — are held per the manufacturer's wear-parts list, sized by your duty cycle: a training platform running eight hours a day eats actuators faster than a research rig running eight hours a month. Insurance parts — a complete spare actuator or drive unit — are justified when downtime costs more than the part, which for commercial simulators is nearly always. Obsolescence is the slow risk: control electronics age out of support before steel does. At purchase, ask for the control platform's support horizon and the migration path when it ends; a Stewart platform's mechanics will outlive two generations of its controllers, and the second controller should be a planned purchase, not a crisis.

When to call the manufacturer

Escalate early on three symptoms: repeatability that drifts between calibrations (something is wearing progressively), response delay that grows (control or drive degradation), and any structural finding — cracks, fretting, fastener elongation — anywhere in the load path. The first two are economics; the third is safety, and it grounds the platform until assessed. The manufacturer's stated 24-hour after-sales response exists for exactly these calls; use it before the platform makes the decision for you.

Condition monitoring: let the platform report itself

The platforms' closed-loop instrumentation doubles as a condition-monitoring system if anyone reads it. Trend the channels that predict wear: actuator current at a reference move, following error at a reference pose, hydraulic pressure ripple where fitted, and temperature at duty. A platform whose reference-move current climbs month over month is announcing a future failure at a time of its choosing; read monthly, and the announcement becomes a planned actuator swap during scheduled downtime instead. The data is already being generated by the control system — the maintenance programme's job is to look at it on a rhythm, with thresholds agreed while the platform is healthy.