Rotary Indexing Table Selection in Thailand: RFP and FAT/SAT

When planning an assembly or inspection machine for a factory in Thailand, it is tempting to start by asking how many stations the rotary indexing table should have. Rotary indexing table selection needs a broader specification: process sequence, indexing and dwell times, combined inertia of the dial, fixtures and workpieces, eccentric loads, process forces, rotary utilities and safety functions. This guide takes purchasing and engineering teams from initial sizing through an RFP and FAT/SAT using the same set of assumptions.

It applies both to new machines and modifications to an existing line. Even when product geometry and target takt may change, listing the open assumptions helps suppliers provide quotations that can be compared.

Start rotary indexing table selection with the process map

A rotary indexing table moves a dial carrying workpieces through defined angles. Feeding, location, pressing, screwdriving, inspection and discharge can take place around it while the dial is stopped. Eight operations do not automatically mean eight index positions. Can loading and unloading share a location? Does inspection require another image? Where will rejected parts leave? Is a spare station needed for a future process? The answers change the station count. Also separate operating stations from the access space required for maintenance.

For each station, write down the operation, longest processing time, state of the workpiece clamp and frequency of fixture changes. Suppose, purely as a planning example, feeding takes 0.8 seconds, pressing 1.4 seconds, image inspection 1.1 seconds and unloading 0.6 seconds. The longest operation that cannot run during indexing, pressing, is a candidate lower bound for dwell. Clamp confirmation, sensor decisions, PLC handshakes and safety checks add time. Measure real operations and agree with the builder which actions can truly run in parallel.

Station count also affects index angle, available floor space, fixture diameter and operator reach. More positions make each angular move smaller but can crowd process heads and guards. Prepare both a station process map and an interference drawing before comparing models. A catalogue’s maximum number of positions should never be the sole reason for choosing it.

Specify indexing time and dwell time separately

At a conceptual level, cycle time comprises indexing plus settling and the work performed during dwell. When feed, processing and inspection run concurrently, dwell is governed by the longest parallel path, not the sum of all tasks. A process that must wait for a clamp adds that clamp time in series. Ask when the “index complete” signal becomes stable under the largest workpiece, highest process force and frequent restart conditions, rather than quoting an average takt alone.

“Rotation in 0.5 seconds” is ambiguous. Does the interval end on arrival, after settling, or after position confirmation? What vibration is acceptable at the process point? Shortening dwell on paper can cause camera blur, misaligned screwdriving or movement during pressing. An excessive dwell assumption can also distort a comparison of drives and sizes. Put target takt, measurement start and end points, and test load in the RFP together.

For a cam indexer, changing input motor speed alone does not guarantee the desired indexing-to-dwell ratio. Sankyo’s SANDEX selection material works through number of stops, indexing angle, input speed, inertia of the dial/fixtures/workpieces and required torque. It also discusses control of the input side when a long dwell ratio is needed. Request a documented calculation of cam curve, input speed, index angle, dwell ratio and output behaviour. For a servo unit, verify the motion profile and settling under the stated load.

Calculate the combined inertia of dial, fixtures and workpieces

Keeping total mass under the payload limit is not sufficient. Moving the same mass farther from the rotation centre increases its moment of inertia and changes acceleration torque and settling. Record the mass and centroid radius of the dial, fixtures, clamps, workpieces, fasteners, and rotating cables and tubes. If a later product may require heavier fixtures, calculate that maximum configuration as a separate case.

For an estimate, add each component’s inertia about the table axis. When the component’s own inertia is material, use the parallel-axis theorem: inertia about its centre of mass plus mass multiplied by the square of the distance to the table axis. An ideal solid-disc approximation is useful for an early estimate, but a dial with cut-outs or asymmetric fixtures calls for CAD mass properties. State assumptions, units, maximum configuration and manufacturing tolerances, and compare values at the same axis and operating conditions as the supplier’s allowable inertia.

Explicitly hypothetical sizing example

This is an illustrative assumption, not a specification or guarantee for any product. Assume a uniform thin circular dial weighing 20 kg with radius 0.30 m. Its central inertia is J = 1/2 × m × r² = 0.90 kg·m². Put four fixtures of 2 kg each at radius 0.25 m; treating each as a point mass gives 4 × 2 × 0.25² = 0.50 kg·m². Add four workpieces of 0.5 kg at that radius: 4 × 0.5 × 0.25² = 0.125 kg·m². The simple total is 1.525 kg·m².

This total cannot approve a table. Fixture self-inertia, clamps, actual mass distribution, bearing support, eccentricity, process reaction force, acceleration profile, duty and life have not been included. If some stations are empty during the cycle, assess that eccentric state separately from the total mass. Supply the assumptions and CAD to the manufacturer and request a selection calculation with stated margins.

Separate eccentric, axial, radial and overturning loads

“Allowable payload” is not a single rating that covers every force. Workpieces concentrated on one side change eccentric torque and bearing moment. Gravity acts differently in horizontal and vertical installations. A press or screwdriver applies force during dwell in addition to static workpiece weight. A force applied away from the table centre creates an overturning moment.

At minimum, specify axial load, radial load, overturning moment, eccentric load, processing torque and possible impact separately. Conditions differ between motion and dwell, clamped and unclamped states, and normal running and an abnormal stop. Do not take several isolated catalogue maxima and assume they are available simultaneously. Sankyo’s RollerDrive material shows that support tables and clamps affect permissible conditions. Map each manufacturer’s definitions to the actual load cases in its selection calculation.

For pressing, welding, drilling or staking on the dial, design the load path for the whole machine. If a secondary support or mechanical clamp closes that path, evaluate its position accuracy, maintenance and interference as well. Decide whether the index mechanism carries process force or each station diverts it; this choice must be clear in the machine builder’s responsibility scope.

Compare cam and servo indexing by expected changes

A cam unit mechanically repeats a designed index pattern and is a candidate for fixed stations and repeated cycles. A servo unit can program positions and motion profiles, making it a candidate when products require different angles or moves. Neither “cam is always faster” nor “servo is always more accurate” holds as a general rule. Series, load, gearbox, control, settling and stiffness all affect the result.

Compare current and future station counts, whether variable indexing is needed, changeover frequency, target takt, load variation, positioning and repeatability needs, holding after power loss, and the controls that maintenance staff can support. For a highly repetitive fixed process, a mechanical index proposal deserves consideration. For frequent product changes, value the flexibility explicitly. Programmable motion still requires renewed checks of the safe operating envelope and fixture interference after changes.

DESTACO presents mechanical indexers among its rotary positioning families, while WEISS offers rotary tables and related controls. Product pages help identify candidates; a performance percentage for a specific model under specific conditions is not a claim about an entire drive class. If a quotation uses a performance figure, tie it to model, load, test conditions and manufacturer documentation.

Plan rotary utilities and safety from the outset

Air cylinders, vacuum grippers, sensors, lights and powered clamps on a moving dial require air, power or signals across the rotation. Pneumatic rotary unions, electrical slip rings and cable carriers with limited rotation all have different maintenance needs. Check not only port count but also pressure, flow, vacuum leakage, current, signal integrity, exposure to oil and dust, angle of travel and replacement access. More components may constrain the central bore and shaft area.

Define the interface between the local machine builder and table manufacturer. Who selects the union and slip ring, routes rotating wiring and stocks consumables? If air leaks or a wire breaks, which signal triggers a safe stop and how is a held workpiece recovered? In a Thai factory, imported-part lead time and local maintainability matter in the quotation. A spare-parts list, drawings, terminal schedule, PLC I/O list and replacement instructions in an agreed language clarify responsibility after start-up.

For safety, identify hazards and intended use before carrying out a risk assessment. ISO 12100 sets out principles for machinery risk assessment and risk reduction; ISO 13849-1:2023 addresses design and integration of safety-related control parts. Neither means that buying a particular table makes a machine safe. Evaluate access to rotation, entanglement, ejected workpieces, restart when a door opens, retained air pressure, power loss and maintenance mode across the whole machine. Design guards, interlocks, emergency stops, safe stop/hold behaviour and recovery from that assessment, then verify them on site.

What to include in a rotary indexing table RFP

An RFP must align assumptions, not merely name a preferred model. Provide drawings and mass ranges for workpieces, process layout, stops and angles, target takt, station dwell operations, dial and fixture CAD, maximum eccentric case, process forces and their points of application, installation orientation, utilities, environment, guarding and local power/air conditions. Define FAT/SAT acceptance and deliverables before purchase. Ask suppliers to list all assumptions and update them under document control once questions are answered.

Pair requirements with evidence so that different prices remain comparable.

Compare fixture, guard, installation, commissioning, training, spare parts and future product-change costs separately from machine price. Align currency, taxes, shipping terms, warranty start date and local service response. A cheaper bare table may become the more expensive machine once unions and safety circuits are added. Our guide to choosing an automation equipment manufacturer covers scope and service responsibilities. The Thai factory ultrasonic welding equipment selection guide likewise shows why process and acceptance conditions should be fixed before comparing quotations.

Inspect the site before comparing offers. In addition to floor space and delivery route, measure the opening direction of adjacent machine doors, workpiece transfer height and space needed to remove a motor or union for service. Confirm site voltage, grounding, air pressure and quality, and any cooling or exhaust needs. When these are unknown, state them as exclusions and define the conditions for a revised offer after a site survey.

Scope boundaries matter especially during installation and commissioning in Thailand. An imported table may meet its specification while a locally built fixture weighs more than planned. Diagnosing the problem requires documentation from both suppliers. List what the machine builder guarantees for the complete machine, what the table manufacturer guarantees as a component and what utilities the plant provides. Even if one contact owns the inquiry, define the route for engineering investigation and repair.

Ask suppliers questions that demand a calculation

“Can you handle this?” invites a broad yes. Ask which model meets the specified maximum inertia, minimum indexing time and maximum eccentricity in the same operating case, and which ratings were checked. Ask for a drawing showing whether the stop-position holding system can take process force or needs an additional clamp. Define changes that would trigger reselection and the limits on future fixture weight.

Catalogue accuracy can be compared with process needs only after checking measurement position and load. Table index accuracy does not directly guarantee hole location on the workpiece: fixture, work holding, frame, temperature, tool and sensor errors add up. Judge final acceptance by measurements on the workpiece and machine while running the process. Agree in advance who supplies gauges, which workpieces are tested and how many measurements are required.

Define accuracy terms through their measurement methods

Indexing accuracy, repeatability and stiffness during dwell describe different properties. Specify where, at what load and temperature to measure the difference between commanded and actual stop angle, the scatter after repeated returns to one position and displacement under process force. The same angular error creates a different workpiece displacement near the centre than at the dial perimeter. Decide whether the operation needs angular repeatability or relative tool-to-part location before comparing a performance table.

For inspection, determine whether image capture can begin as soon as the index-complete signal appears. Mechanical stop and the disappearance of image blur may occur at different times. Make a timing chart that includes exposure, lighting, vibration settling and confirmation of workpiece retention, then measure through to the pass/fail decision. For pressing, measure tool-to-workpiece position while the process force acts. No-load repeatability alone does not establish process quality.

If products will change, list tolerances by product. A longer fixture or heavier workpiece can enlarge displacement at the outer stations even if the first product has ample margin. Document why the most demanding product represents the family; test more than one product if it does not. This connects FAT/SAT results with production quality control. Keep raw data by condition, not only averages, together with instrument calibration records. Reproducible methods help determine whether a later position problem reflects deterioration or a workpiece or fixture change. Acceptance tests also establish maintenance baselines.

Make FAT/SAT an acceptance test of the process

Keep purchasing, engineering and maintenance on one revision

Before ordering, check that all three teams use the same specification revision. Engineering confirms the maximum workpiece and fixture case used in calculations; purchasing checks the quoted model and options; maintenance checks stop/recovery and consumables. If dial diameter changes but an old inertia calculation remains in the offer, no one can tell whether the purchased drive suits the changed design. Put drawing number and revision on the selection calculation, offer and FAT protocol.

At the quote review, compare exclusions before declaring one vendor cheaper. List the table, motor, drive, home sensor, support, clamp, rotary union, slip ring, guard, installation, testing and training. Verify assumed plant power and air with measurements. Once specification differences are separated from commercial differences, the team can debate priorities rather than only negotiate price.

For spare parts, ask about lead time and substitutes as well as price. A joint seal or sensor held in Thailand supports a different recovery plan from an imported item. Who saves drive parameters and restores them after replacement? A drawing alone may not explain mechanical alignment or required tools. Ask the local maintenance team to demonstrate operation and fault recovery during commissioning and confirm the instructions suit its language and skills.

Define when changes require reselection

Product launches often change workpiece shape, fixture, inspection camera or feed direction. Set change criteria early. Update affected load cases if fixture mass or radius, workpiece distribution, indexing time, dwell force, rotating electrical load or operating frequency changes. “Only a little heavier” is not an engineering criterion. Compare CAD mass properties and process timing before and after the change and ask the supplier whether the original calculation remains valid.

Unlimited margin can make the first machine unnecessarily large and expensive. State realistic plans such as one additional product or a specified fixture envelope, then decide where to reserve capacity. Comparing upfront and later modification costs allows the business team to value flexible indexing or a larger model.

Put test scenarios in the acceptance document

Examples include running at target production speed with maximum workpieces at every station; repeatedly stopping and restarting with one position empty and thus eccentric; and measuring location while the press applies its specified maximum force. These illustrate how to describe tests, not universal durations or tolerances. Derive test time, repeat count and limits from the actual machine. Record workpiece ID, fixture revision, sensor settings, instruments and where data will be stored so another tester can reproduce the result.

If FAT uses substitute workpieces, equal mass alone may not be enough: a different centroid radius changes inertia. Demonstrate equivalence by calculation and carry tests with real parts into SAT. If the support and surrounding machines are installed only in Thailand, make precision and interference checks there mandatory. Split tests without losing continuity of responsibility.

Do not define the continuous-run acceptance criterion as “the target takt was achieved once.” Record which product and fixture ran, how many cycles were completed, and the stop and restart conditions. Capture indexing time, settling delay, process completion time and error count separately for each cycle. This can expose a slow station even when the average looks good. Derive test duration and acceptable counts from the actual product and machine requirements; there is no universal value here. Synchronise measurement logs with video timestamps so that a later site problem can be compared with the acceptance run.

Recovery after a stop also belongs in acceptance testing. If an emergency stop occurs while a part remains clamped, decide what happens to air pressure before opening the guard, whether homing is required on restart, and whether the in-process part is reworked or rejected. Separate the cause shown on the HMI, checks for the operator and functions reserved for maintenance. Test a procedure that prevents the same part being pressed or processed twice by mistake. Have site staff carry out recovery on the actual machine, then record the time required and screens that cause confusion.

During product changeover, verify that fixture identification, recipe, index angle, camera settings and process-head settings all belong to the same product. Decide how the first part after changeover is inspected and approved before production resumes. Correct table positioning cannot protect quality if the wrong fixture or an old inspection recipe is active. Run changeover scenarios at both FAT and SAT and keep operation records in a language that local staff can use.

FAT, at the builder’s facility before shipment, checks index positions, cycle, abnormal stops, guard doors, air-pressure loss and homing with actual or agreed substitute workpieces. SAT checks the same scenarios after connection to Thai plant utilities and upstream/downstream machines. Distinguish conditions. A video of free running or a no-load cycle time does not establish acceptance under maximum production load.

For each test list conditions, steps, measurement, pass criterion, record and corrective owner. Include the heaviest and most eccentric fixture case, the longest dwell process and recipe/fixture matching after changeover. Test safety functions against the risk assessment: door and emergency-stop activation, stopping behaviour, prevention of unexpected restart, service mode and recovery. The machine designer and safety lead determine applicable standards, required performance and validation.

Documents are acceptance items too: final electrical and pneumatic drawings, PLC program and backup method, parameter list, fixture drawings, bill of materials, lubrication and replacement intervals, diagnostics, spares and local contacts. If a model changes, align revisions of calculations, drawings, tests and manuals. Record the owner, deadline and retest conditions for any conditional acceptance.

Frequently asked questions

Can rotary indexing table selection be based only on the number of stations?

No. The number of stations follows from the process map. Compare models only after specifying timing, combined inertia, eccentricity, process reaction forces, support, rotating utilities, safety and maintenance. Start with the station map and maximum-load case for the supplier’s selection calculation.

Is a cam or a servo indexer faster?

There is no universal answer. Compare proposals using the same workpiece, fixture, index angle, positioning and settling criteria, and duty. Include future changeovers and variable moves. Do not interpret a performance claim for one product series as a property of an entire drive method.

What drawings are essential in an index table RFP?

Start with workpiece and fixture drawings, process layout, CAD mass properties for the dial, points of process-force application, and interference drawings for guards and adjacent equipment. If details are missing, state assumptions and update dates, and agree what must be recalculated after a change.

How should rotary table FAT and SAT be divided?

FAT checks the built machine before shipment. SAT confirms site utilities, linked processes and actual operator use. Some scenarios need repeating in both places. Agree on parts, conditions, measuring tools and acceptance criteria before purchase.

Conclusion

Select the rotary indexing table by deriving positions and real takt from the process map, calculating combined dial/fixture/workpiece inertia and load cases, and comparing cam and servo options against likely changes. Put utilities, safety and maintenance into an RFP and verify the complete process through FAT/SAT. Use selection calculations and test results under the same conditions, not a single catalogue number, to choose a machine that can be compared fairly and maintained in service.

If your Thai factory is preparing an automated-machine project, contact TOMAS TECH even while the process map or maximum-load cases are being assembled. We can help align responsibility for the table and surrounding equipment.

References

• Sankyo Seisakusho, SANDEX Alpha series catalogue — selection sequence for stops, index time, inertia and torque.
• Sankyo Seisakusho, RollerDrive rotary table selection guide — loads, inertia, process force and installation.
• Sankyo Seisakusho, RSR catalogue — support table and allowable-load conditions.
• DESTACO, Rotary Positioning — positioning product families.
• WEISS, Download Center — rotary tables and controls.
• ISO 12100:2010 — machinery risk assessment and reduction.
• ISO 13849-1:2023 — safety-related control system design.

*Note: The calculation is hypothetical. Final model selection, safety design and acceptance criteria must reflect current manufacturer information, the machine risk assessment and site conditions.*