When selecting an ultrasonic plastic welding machine, the first things to determine are not frequency or output. Rather, you must first consider the resin to be joined, the joint configuration, the required quality, the production volume, and who will supply the workpieces to the equipment. For those responsible for launching resin case and sensor housing joining processes at Thai factories, this article organizes the practical steps from confirming joinability, to equipment configuration, RFP, and FAT/SAT. Please note that numerical values must be determined experimentally according to the application, and the examples in this article should not be treated as guaranteed values.
Selection of Ultrasonic Plastic Welding Machines Begins with “Establishing Joinability”
In ultrasonic welding, vibration from the horn surface is transmitted through the parts to the joint interface, where heat is generated, melting the resin, and after cooling, a joint is formed. Branson’s explanation also describes this principle. Therefore, it is not possible to select equipment based on the idea that “this resin will always melt with a 20 kHz machine.” Even with the same resin name, changes in grade, additives, molded part rigidity, distance to the joint line, and the method of supporting the parts will alter the required amplitude and the results. Branson’s explanation of the principle and Dukane’s selection procedure both indicate the necessity of evaluating parts and joining conditions before specifying equipment.
For example, in the process of closing a two-part resin case, the evaluation differs between joints for appearance only and those requiring airtightness. For the former, you may focus on the continuity of the joint line, burrs, and surface scratches. For the latter, you need methods for leak testing, pressure conditions, holding time, and pass/fail criteria. If circuit boards or filters are built in, the effect of vibration on internal components is also an evaluation item. After joining, confirm with the finished product to avoid situations where “the appearance is good, but internal components are damaged.”
The first documents to provide to the equipment supplier are not specifications for a model, but rather part drawings, the official resin name and grade, the range of molding conditions, prototypes, and a list of required quality characteristics. Do not decide joinability based solely on the material manufacturer’s data sheet; include a stage for testing with actual workpieces in your procurement plan. Also indicate the priority of required quality characteristics. It is not always possible to maximize strength, airtightness, appearance, dimensions, and electrical properties simultaneously, and processing conditions will change depending on which items are essential for customer approval.
Determine Feasibility First Based on Resin, Joint, and Workpiece Support
Do Not Decide Based Only on Resin Name; Confirm the Combination of Parts
Dukane’s procedure for the initial stage of ultrasonic joining involves confirming whether the target is a thermoplastic resin, whether there is compatibility in the case of dissimilar resins, and whether the design allows ultrasonic energy to reach the joint. In practice, you should prepare not only the resin abbreviation from the material sheet, but also the grade, fillers, colorants, recycled material ratio, and any planned changes in suppliers. If there is significant variation in mass production lots, it is not sufficient to confirm that conditions were met in prototyping alone. Use the material and molding dimensional variations that may occur in mass production to check the process window.
Even if the joining partner is the same resin, if there is a large difference in wall thickness, the energy transfer from the horn contact point to the joint interface will be uneven. Thin walls or protrusions may deform first. If the molded part is warped, the mating surfaces may float when placed in the jig, resulting in a mixture of excessive melting and incomplete welding in localized areas. Provide the supplier not only with good parts, but also with actual parts close to the upper and lower dimensional limits, and include them in the test range.
Include Joint Geometry and Energy Directors in the Design Scope
Details of the joint such as energy directors are important for initiating melting at the intended location. For parts where the mold has already been finalized, there is limited room for modifying the joint. Therefore, conduct tests separately for “whether joining is possible with existing parts” and “recommended proposals when mold changes are involved.” Note that in equipment quotations, the costs for these design changes and prototyping periods are often omitted.
Specify the position of the joint line, the required penetration depth, flash escape, permissible deformation on appearance surfaces, and positioning references in the drawings. If numerical values are not yet finalized but are included in the RFP, clearly state that these are ‘items to be agreed upon through testing.’ This is to prevent them from being used unilaterally as expected values during acceptance. In addition to photographs, determine the necessary methods for the product, such as cross-sectional observation, destructive testing, or leak testing.
View the contact surfaces of the lower jig and horn as integrated with the product side
If the surface pressed by the horn and the surface supported by the lower jig are misaligned, the part will tilt, causing the load to concentrate on only a portion of the joint line. For workpieces with complex curved surfaces, simple flat supports do not provide reproducibility. It is necessary to design jigs that do not damage the workpiece, can withstand vibration and force during welding, and do not catch when removing the part. Jigs are also subject to wear, cleaning, and replacement, so include spare parts and change history in the equipment specifications.
The horn is not merely a pressing jig, but a part of the acoustic stack that includes the transducer and booster. Dukane explains that the amplitude at the horn’s working surface is determined by the gain of each stack element. Insufficient amplitude can lead to weak joints or incomplete melting, while excessive amplitude may cause process instability. Do not select the required amplitude based solely on a single catalog value; confirm it with the actual material and workpiece. Dukane’s explanation of amplitude.
For general considerations in jig design, please also refer to our article on jig design and manufacturing. However, in ultrasonic joining, in addition to positioning, also evaluate the interaction between the acoustic stack and part support.
Equipment Configurations Compared When Introducing Ultrasonic Welders
Estimate Separately for Manual Presses, Semi-Automatic Cells, and Fully Automatic Cells
After confirming the feasibility of joining, select the method for supplying and removing parts. For prototyping or high-mix, low-volume production, it is often easier to start with a press type where the operator sets the parts. On the other hand, if there are high requirements for takt time and quality recording, consider a cell that integrates supply, orientation verification, clamping, welding, inspection, and discharge. Dukane’s selection procedure also describes the use of portable probes, presses, rotary indexers, and dedicated automation as appropriate. Compare all methods, including the actual required takt time and changeover.
For example, even if you compare only the ‘welding time per piece,’ adding the time for part setting, presence detection, post-weld holding, removal, and inspection may result in not meeting the line takt time. Whether to use parallel stations, a rotary system, or treat the welding process as a bottleneck should be determined using a cycle diagram of the entire process. In cells with manual part supply, also measure the operator’s hands-off time and arrange jigs so that operators are not forced to align parts in awkward positions with one hand.
Even when purchasing a standard machine from an equipment manufacturer, the horn, lower jig, guard, workpiece supply, PLC, and result data linkage may fall under separate scopes. Conversely, when placing a bulk order with an SIer, confirm the boundary between the manufacturer’s warranty for the welding machine itself and the warranty for the entire cell. Our article on special-purpose machine design and manufacturing covers general ordering scope organization. This article focuses specifically on resin joining recipes, acoustic stacks, and quality judgment within that context.
Narrow Down Frequency, Output, and Drive Method Based on Application Testing
The product information for the Branson 2000Xd lists frequency options of 15, 20, 30, and 40 kHz, as well as welding modes such as time, energy, distance, and peak power. However, these are specifications for the specific model and are not recommended values for all resins and dimensions. While using the product specifications as an example for functional comparison, request that suppliers present candidates and their rationale based on prototyping results.
Avoid the simplistic assumption that higher frequency and lower output are always advantageous for delicate parts, or that higher output is always sufficient for large parts. The results are determined by the combination of horn shape, amplitude, load, actuator rigidity, and workpiece support. When comparing pneumatic and servo drives, consider not only the superiority of specifications, but also the required load and speed profiles, monitoring granularity, maintenance system, and local parts supply.
Link Welding Modes and Process Data to Acceptance Criteria
In time mode, even if the time is fixed, differences in results due to material variations will remain. Modes such as energy, distance, and peak power do not guarantee quality based solely on the set values. Correlations between quality characteristics and process parameters should be verified through prototyping, and upper and lower monitoring limits for results, as well as their correlation with product testing, should be established. Equipment with multiple modes and graph monitoring functions, such as the 2000Xd, is useful for examining these correlations, but the data format and output conditions must be confirmed at the time of order placement.
If it is necessary to link product serial numbers with welding results, specifications must include communication with PLC or higher-level systems, clock synchronization, retransmission in case of data loss, and isolation of NG (non-conforming) workpieces. Having only numerical logs does not complete traceability. The design must detect barcode misreads and duplicate scans, and prevent unsaved data from being sent to the next process. The communication methods supported by the equipment model must be matched with the IT/OT requirements on the factory side.
Required Items to Include in the RFP for Resin Welding Equipment
An RFP that simply states “one set of ultrasonic welding machine, 1 unit” is too vague. The purpose, workpiece, quality, capacity, interface, working conditions, and acceptance method should all be compiled in the same document so that suppliers can provide quotations based on the same scope. Any undecided conditions should not be concealed but clearly state who will determine them and when through testing. In the quotation comparison table, list standard machines, horns, lower jigs, guards, automatic supply, inspection, data, installation, training, and spare parts as independent items.
| RFP Item | Example Description / Confirmation Points | Evidence Required for Acceptance |
|---|---|---|
| Workpiece | Drawing revision, resin grade, molding lot, built-in components | Identification records of test subjects |
| Joint Quality | Strength, leak, appearance, dimensions, impact on internal components | Agreed test methods and raw data |
| Mass Production Capacity | Target takt time, operating hours, frequency of mold change, supply method | Time records of continuous operation |
| Process Control | Recipe, change authority, monitoring values, alarms, NG isolation | Demo including abnormal input |
| Jigs & Horns | Drawings, material, replacement method, spare parts, maintenance | Parts list and replacement demonstration |
| Interface | Power supply, compressed air, PLC, data format, clock synchronization | I/O check and data verification |
| Safety & Maintenance | Guard, emergency stop, cleaning, noise check | Risk assessment and maintenance procedures |
| Scope of Delivery | Installation, Thai language training, warranty, local support | Signed handover documents |
Quality standards such as “no leakage” or “strong” are not sufficient for acceptance. If customer specifications exist, reference their test jigs, conditions, sample quantities, and pass/fail criteria. If customer specifications are not yet determined, establish standards based on DOE or prototype results, and have both parties approve them in writing before FAT. In process window testing, include not only nominal dimensions but also parts close to the tolerance limits. It is also important not to treat prototypes and mass production mold products as equivalent.
Questions to suppliers should not be limited to “Is it possible?” but should specify “With which test, which quality characteristics, and how many samples will be verified?” Additionally, agree on the cost of workpieces lost in destructive testing, the testing institution, responsibility for mold modifications, and retesting costs. Branson’s application support clearly defines the stages of initial review, joint and equipment design, testing, and implementation training, serving as a reference for including the prototyping process before equipment selection in the procurement schedule. Application Support.
Do Not End Supplier Comparison with Only Price and Catalog Tables
In the comparison table, list the basis for joint formation, design responsibility for horns and jigs, width of the process window, ease of extracting quality data, and specific conditions for local service support. Even if “support available” is written, the strength of the basis differs depending on whether the evaluation workpiece is actual resin or a substitute material. When borrowing samples for testing, separate and test good products, products with molding variations, products with different assembly orientations, and products with built-in components. Test results should be saved not only with set values but also linked to workpiece lots and results.
Items to be included in the estimated cost are not limited to the main equipment. Separate costs for redesigning dedicated horns, spare jigs, guards, exhaust and soundproofing needs, mold change parts, operator training, maintenance training, annual calibration, local installation, and import and startup. Differences in quotations based on parts or foreign currency should also be compared under unified contract conditions. For specific investment recovery, calculate using your own good product rate, takt time, and personnel allocation along with the quotation, and do not reuse assumed values from general articles.
Sequence of Confirmations When Transitioning from Prototyping to Mass Production Conditions
In prototyping, first search for combinations that can be joined safely, and then search for the range of conditions that can be maintained in mass production. Making the strongest test piece in a single trial and consistently producing acceptable products even when molding lots or operators change are different matters. In initial tests, fix the resin, joint shape, horn surface, and lower jig, and vary the main welding conditions to observe trends. For each test, record the history and conditions of the parts, and cross-check with product tests such as airtightness and strength. If results vary, before continuing to fine-tune only the set values, investigate molding dimensions and the seating condition on the jig.
Once candidate conditions are identified, introduce the variations expected in mass production. For example, multiple molding lots, upper and lower dimensional limits, after mold changes, after equipment restarts following shutdowns, operator changes, and after jig replacements. These are not “tests to increase the number of tests,” but actual situations where conditions change. If products pass only at the center of the mass production conditions and defects occur at the extremes, decide whether to improve the equipment parameters, product design, molding control, or jig structure. If the acceptance window remains narrow and the equipment is incorporated into an expensive automatic machine, the equipment may simply produce defective products faster.
The test plan should indicate, in a simple table, which factors are to be fixed and which are to be varied. If you compare only amplitude and load without recording the material lot and molding dimensions, you may misattribute the cause of differences. Conversely, if you change all factors at once, it becomes difficult to reproduce the results. If necessary, use design of experiments (DOE), but before preparing formulas for analysis, first establish jigs and measurement procedures that can accurately assess quality characteristics. This is because if the specimen cutting position, leak test sealing method, or holding method for destructive testing changes, the differences in measurement methods may outweigh the differences in joining conditions.
Example of a Sample Request Form to Attach to the RFP
The sample request form should specify the part number, drawing revision, mold to be used, resin grade, color, lot, molding date, storage conditions, quantity, and protection method during transportation. It should also clearly state whether there are any built-in components and whether those components can be returned after testing. Quantities used for destructive testing or cross-sectional observation will not be returned, so avoid sending limited pre-production prototypes without a plan. If parts are to be shipped from the Thai factory to an overseas equipment manufacturer, include the transportation period and customs clearance estimates in the schedule.
It is fundamental that the drawing revision and the actual sample match, but during initial studies, samples may be sent without matching the latest revision. In such cases, list which revisions are reflected in the actual sample and which points are not yet reflected. Even if the test report shows a “pass,” if the test was conducted with joining details not present in mass production mold parts, it cannot serve as evidence for mass production approval. Simply including the same part number, revision, and lot on both the sample request form and the test report can reduce misunderstandings in decision-making.
Request the equipment manufacturer to return samples that failed joining, or at least report them with photographs and data. If you can identify the location of the defect, horn marks, flash, fracture surface, and process waveform, you can proceed to investigate whether the cause lies in resin compatibility, joint design, jig support, or parameters. If you only receive a report stating “the test failed,” you cannot determine whether to change the model or the part. Receiving results in a format that can be used for the next design change is a condition for making effective use of prototype costs.
After prototype approval, determine the point at which design changes to molds and materials are frozen. If the part changes, the horn contact surface, lower jig support, and required melt depth may also change. If changes are unavoidable, record which tests need to be re-conducted on a change management sheet. Before evaluating process capability, it is also important to check the calibration of measuring instruments and the variation between measurers. If measurement fluctuations are mistaken for equipment instability, it can lead to unnecessary equipment modifications or excessive re-prototyping.
The contract for equipment introduction should also specify how to respond if the required quality for mass production is not achieved under the conditions confirmed during prototyping. Decide who will investigate causes attributable to parts, jigs, or settings, which documents will be shared, and how many times retesting will be included. If you only broadly assign responsibility to the supplier, progress will halt when a part design change becomes necessary. It is practical for the product design, molding, quality, and equipment representatives of the purchaser to participate in the same decision meeting and determine the next actions based on the results.
Once the process window has been confirmed, separately specify the standard conditions and the allowable range for changes. If operators are allowed to freely adjust settings on the grounds of daily variation, the evidence from prototyping will no longer be able to explain the mass production conditions. Define the person responsible for approving changes, the number of confirmations required after changes, and where to store the change history. When handling multiple product types on a single machine, consider a system where reading the part number displays the corresponding recipe and jig number, and the machine will not start with an incorrect combination. Even just aligning the part number notation between production management and the equipment side can reduce confusion during on-site commissioning.
If the evaluation jig used in prototyping differs from the mass production jig, record those differences. For example, if parts were held by hand by engineers during evaluation but are automatically clamped in mass production, the way force is applied and seating may change. Retest using the mass production jig and supply method, and do not approve based only on a short demonstration video; retain the results of the specified quality tests. Also, confirm in mass production operation whether horn marks appear on the product’s visible surfaces or whether the still-warm joint area is deformed during removal.
How to Inspect Ultrasonic Welding Cells in FAT/SAT
FAT is conducted at the supplier’s site before shipment, and SAT is conducted in the actual environment after installation at the Thai factory, verifying pre-agreed items. Rather than repeating the same checks at both stages, FAT confirms equipment functions, quality with test workpieces, alarms, and data, while SAT checks whether these can be reproduced under local power supply, compressed air, personnel, upper-level systems, and actual material lots. In both cases, document pass/fail criteria in advance, and for items that cannot be judged, record them as outstanding issues with the responsible person and deadline.
The inspection plan should list, line by line, the test number, prerequisites, part number and lot of the workpieces to be used, equipment recipe, person conducting the test, expected results, actual measured results, and storage location of evidence files. If you simply write “normal operation confirmed,” it will be impossible to trace what was checked later. Before starting each test, record the software version, PLC program version, and identification numbers for the horn and jig; if modifications are made during testing, determine the scope of retesting. If defects found in FAT are to be corrected on-site, agree in advance on who will make the corrections and which tests will be redone before SAT.
The scope of inspection includes not only the equipment itself but also the flow of workpieces. For example, if two types of cases come from the previous process and the design selects the recipe by barcode, check that the equipment does not operate when the wrong part number is entered, and that it is not possible to manually select an incorrect recipe if barcode reading fails. If the subsequent process receives the product’s pass/fail status, test the timing of NG workpiece ejection and ensure that unprocessed items are not treated as good products after a power outage or communication failure. Such boundary conditions are not described in the welding machine’s main catalog, but they are directly linked to the quality of the mass production cell.
In FAT, Insert Abnormalities in Addition to Continuous Operation of Good Products
Testing only with a certain number of good products flowing through the system will overlook mass production issues such as reversed workpieces, missing components, unmounted jigs, incorrect recipes, pneumatic pressure drops, data communication interruptions, and recovery from emergency stops. Abnormality injection testing should be planned within a range that does not place the equipment in a dangerous state, and interlocks, NG (non-good) isolation, and recovery procedures should be documented. It is also necessary to check for prevention of contact between the horn and jig, operation when the safety guard is open, and recipe change authorization.
Joint quality should not be judged solely by appearance; instead, measure the strength, leakage, dimensions, and internal component functionality as agreed upon in the RFP. Include tests of parts produced at the upper and lower limits of mass production conditions, and require delivery of the raw data and configuration files. While small variations in welding time or energy are informative, they alone do not guarantee the quality of the final product. Confirm the relationship with quality characteristics.
In SAT, confirm the actual conditions and operations at the Thai factory
In SAT, confirm power quality, pneumatic pressure, temperature and humidity, actual component supply, line stop signals, handshake with the PLC, time synchronization, and data storage locations. Also test whether the local team can execute work procedures, alarm displays, replacement tasks, and daily inspections in Thai. Equipment that only operates when the manufacturer’s engineers are present cannot be transferred to mass production. Confirm that local maintenance personnel understand the procedures for exchanging jigs and horns and can perform initial troubleshooting in the event of an abnormality.
Test samples should use the resin lots and mass production mold parts that will be used locally, and not simply replicate FAT conditions; also check whether transport vibration or realignment after reassembly has any effect. This is why passing FAT is not a substitute for SAT. At the end of SAT, hand over mechanical drawings, electrical drawings, PLC backups, master versions of recipes, maintenance lists, spare parts lists, training records, and a list of unresolved issues.
Anticipate causes of quality deterioration after start-up
Even after equipment acceptance, results can change due to horn wear, fastening condition, jig loosening, changes in molding conditions, changes in resin grade, sensor contamination, or pneumatic pressure fluctuations. Separate items for daily and periodic inspections, and decide who will record what. Dukane’s procedures also list stack, jig, guide, wiring, air filter, and setpoint inspections as maintenance items. However, do not simply adopt their inspection intervals; determine them based on the manual and operating conditions of the adopted model.
At the initial stage of mass production, save both the distribution of process data obtained under standard conditions and the results of product testing together. This serves as a reference for investigating when changes occurred if waveforms or distances change later. The control values used here should not simply use the manufacturer’s initial settings as thresholds, but should be set based on test results of good and defective products. Decide in advance whether to stop the equipment, isolate the affected lot, retest, or investigate the cause if the boundary is exceeded. In shift work at factories, use a concise decision table so that responses do not vary by team.
Simply having spare horns and lower jigs in stock does not guarantee peace of mind. Manage identification numbers, corresponding part numbers, start-of-use dates, reprocessing histories, and storage conditions, and determine the confirmation work and confirmation tests after replacement. Since amplitude or contact condition may change after horn polishing or replacement, inspect according to the equipment manufacturer’s procedures. In multi-product cells, incorporating mechanisms or confirmation procedures in the RFP that prevent operation with incorrect horn and jig combinations can help suppress misuse after start-up.
When changing a recipe, record the reason for the change, the target part number, the approver, the test results, and how to revert to the previous version. When collecting process data, also keep a history of updates to NG judgment values. If changing resin suppliers or molds during mass production, incorporate re-evaluation of joint results into change management. Even if the recipe name is the same on the welding machine screen, the process state is not the same if the horn or jig has been replaced. Link equipment part versions with quality records.
Common Misunderstanding: Selection is not the same as for metal robotic welding or servo presses
In metal robotic welding, the focus is on welding current, torch orientation, wire, and tracking of the weld line. In ultrasonic welding of resin, the focus is on energy transfer from the acoustic stack to the joint interface, resin compatibility, joints, horns, and workpiece support. Although both use the terms “welding” or “joining,” the required prototyping and quality testing are different. While automation know-how from robot cells can be applied to transport and safety design, it does not substitute for establishing joining conditions.
In press-fitting with a servo press, the assembly state is judged using load and displacement curves. In ultrasonic welding, load and distance can also be monitored, but the focus is on bonding by melting and cooling at the resin interface. The pass/fail window for press-fitting cannot be directly applied to the welding process. First, distinguish whether the application is joining resin parts, press-fitting metal parts, or insert press-fitting into resin, and design the necessary tests accordingly.
Common Misconception: The Presence of Welding Data Does Not Equate to 100% Quality Assurance
Process waveforms and threshold judgments are useful for detecting and tracking process abnormalities. However, even if the waveform is within the acceptable range, there remains the possibility of material contamination, internal component damage, or invisible leakage. For each critical product characteristic, a combination of process monitoring, destructive sampling tests, non-destructive tests, and final functional tests is employed. Whether 100% testing is required is determined by customer specifications and risk assessment. It is important for both suppliers and purchasers not to equate monitored values with advertised ‘quality assurance.’
Frequently Asked Questions
Should I Simply Choose a 20 kHz Ultrasonic Plastic Welder?
While 20 kHz is a candidate for many applications, the required amplitude, horn, output, and quality will vary depending on the workpiece shape and resin. Dukane explains the application trends for 20 kHz and 40 kHz, but the final decision is made through prototyping with actual parts and testing for the required quality. If you specify only the frequency in the RFP, please include the technical rationale for doing so.
Can Ultrasonic Welding Equipment for Resins Be Used on Existing Molded Parts?
It is possible. However, if the joint detail, mating surfaces, horn contact area, or support on the lower fixture are not suitable, mold modification or an alternative joining method will be necessary. If it is difficult to modify mass production molds, compare the test results using existing parts separately from the cost and lead time for design modification proposals.
What Is the Minimum Required Information for an Ultrasonic Welding RFP?
Part number and drawing revision, resin grade, joint and quality requirements, target cycle time, production volume, supply method, factory interface, and FAT/SAT test methods. For undecided quality values or process windows, clearly state ‘to be determined by prototyping,’ and include prototyping costs and decision deadlines in the estimate.
How Many Samples Should Be Tested in FAT/SAT for Ultrasonic Welding?
There is no fixed number that applies to all products. Decide based on customer specifications, destructive test methods, material and dimensional variation, and the required reliability. Include not only good product samples but also parts near the limits of the acceptable range and intentionally abnormal inputs in the test plan, and record the results.
Summary
Selection of ultrasonic plastic welding machines becomes more concrete when proceeding in the order of resin and joint compatibility, horn and fixture feasibility, correlation between product quality and process data, mass production cell configuration, and FAT/SAT. Manufacturer catalogs serve as reference materials for candidate selection, while prototyping results with actual workpieces and mutually agreed acceptance criteria are the basis for ordering decisions. At the Thai factory, the plan includes re-verification with actual materials and local operation after installation.
If you have drawings of resin parts and target cycle times ready, and are considering whether to order a standalone device or a joining cell, please consult TOMAS TECH. We can help organize the scope to be written in the RFP and the approach to prototyping and acceptance, tailored to your product requirements.
