Wire Harness Crimp Quality Control: CFM, Crimp Height, Traceability and RFP

Introduction: Your Crimp Data Lives Only “Inside the Machine”

This article is written for plant managers, production engineers, quality assurance and maintenance staff in Thailand who are considering a review of wire harness crimp quality control. Crimp height is measured with a micrometer at every setup and written on paper. The fully automatic crimping machines have a crimp force monitor (CFM), but the manual bench presses do not. Shop floors like this are not unusual.

A customer asks, “Please send us the crimp data for the terminals in this lot.” You open the crimping machine screen, only to find that nothing records which wire lot was combined with which terminal reel. Or the data itself has been overwritten inside the machine. EV programs bring more high-voltage large-gauge wire and aluminum wire, and your past experience in setting crimp parameters no longer carries over as it is. If these situations sound familiar, this article is for you.

The scope here is quality monitoring of terminal crimping in wire harnesses, and linking crimp data to wire lots, terminal lots and harness IDs. In the sense of linking process data to individual units, it follows the same pattern as our article on spot welding monitoring. Joining terminals by ultrasonic welding is a different process, so it is only touched on in Section 12.

Here is the conclusion up front. Crimp quality control is not “done once you install a CFM.” CFM, crimp height measurement, cross-section analysis and pull testing each show different things, and none of them replaces another. Before adding equipment, decide these 3 things:

1. Which circuits and which terminals receive focused control
2. What to record for each crimp (or each setup), and how to link it to the wire lot, terminal reel lot, applicator ID, machine ID and harness ID
3. How to record applicator (crimp die) hit counts and first-piece checks at setup

Judging from manufacturers’ announcements, many fully automatic cut-and-crimp machines can handle CFM and measurement data. The starting point is to collect data from your existing machines. Retrofitting CFMs to bench presses should be considered only for critical circuits. Research has also pointed out that CFMs have limitations in the form of false alarms and the effort of re-setting them.

This article is not a recommendation or ranking of specific products. Product features are presented as manufacturers’ announcements. The cost-effectiveness section later in the article is a calculation built on assumed numbers prepared for this article.

Why Harness Plants in Thailand Should Revisit Crimp Quality Control Now

First, let’s look at where Thailand’s automotive industry stands. The figures are given with their source and scope as published; we do not add them together or apply them to any individual company.

• According to a MarkLines article (September 28, 2026) reporting an announcement by the Federation of Thai Industries (FTI), four-wheel vehicle production in August 2026 was 124,646 units, up 10.9% year on year. Growth was led by passenger BEV production (an increase of more than 77%) and pickups for export (an increase of more than 18%). Sales in the same month rose 25.6%.
• According to Kaohoon International, as of October 2, 2026, the FTI is maintaining its full-year 2026 production target of 1.45 million units.
• According to a September 10, 2026 announcement by the Thailand Board of Investment (BOI), electrified and hybrid vehicles took a record 55% of Thailand’s new vehicle market in January to July 2026. This is a share of the new vehicle market (registration basis), not a share of production. The announcement states that combined registrations of BEVs, HEVs and PHEVs exceeded those of internal combustion vehicles for the first time.
• According to the same announcement, cumulative BOI approvals related to the EV ecosystem stood at USD 4.59 billion across 189 projects as of August 31, 2026. It also states that 4 companies (Honda, Mitsubishi, Isuzu and Mazda) will invest a total of USD 1.53 billion to upgrade their Thai assembly lines with robotics and automation.
• A BOI announcement on July 3, 2026 cited EV supply chain investment pledges of more than USD 4.1 billion across 198 projects, and stated that more than 1,200 matchings with more than 800 Thai parts makers had taken place through the BOI’s Sourcing Day. Because the cut-off dates and categories differ from the September announcement, we do not compare the two to calculate an increase or decrease.

On the policy side, the BOI’s September 10, 2026 announcement reports that the National EV Policy Committee has agreed in principle to revise the excise tax structure. The direction is to set 4 tax tiers based on domestic value added and to favor models that achieve deep localization of key electronic components. The more the share of locally made parts is scrutinized, the more parts plants are expected to be able to demonstrate quality through records.

Meanwhile, conditions for parts makers are not easy. According to a Reuters report on May 14, 2026, 10 Thai automotive associations sent a letter to the government stating that parts makers are also losing orders because of zero-tariff imports from China. Krungsri Research (May 7, 2026) projects auto parts production to grow 1.5 to 2.5% per year in 2026 to 2028, and states that Thailand’s auto parts exports in 2025 were USD 24 billion (up 6.9%), the highest in ASEAN. These are all figures for the parts industry as a whole, not figures limited to wire harnesses.

Flooding is also having an impact right now. These are two estimates made at different points in time, so they are listed separately.

• As of October 2, 2026: According to Nation Thailand, Toyota and Honda halted production in Thailand due to flooding caused by more than 320 mm of heavy rain over 3 days. A separate report on the same day said that a levee breached at the Amata City Chonburi industrial estate and that at least 34 factories were directly damaged. The same article put the direct losses to manufacturing at USD 30.4 million (THB 1.018 billion), but this is not the amount for the 34 factories alone; it is a provisional estimate for manufacturing as a whole.
• As of October 8, 2026: According to Nation Thailand, more than 10 parts plants in Zones 7, 8 and 9 of the same industrial estate had halted operations, with a provisional restart planned for October 12. The provisional estimate of industrial losses is about THB 6 billion.

Both are estimates for industry as a whole or for the supply chain as a whole, not losses at individual harness plants. Still, when a stopped line is brought back up, whether you can show through records that you are “crimping under the same conditions as before the stoppage” matters to every plant.

On the materials side, there is movement in aluminum wire. According to an alcircle article dated November 25, 2023, Thai Yazaki Electric Wire plans to expand its aluminum wire facilities at its Suvarnabhumi plant, raising melting capacity from 850 tonnes to 2,500 tonnes per month, and to supply aluminum alloy wire, a material for automotive wire, to the group’s automotive division (completion could not be confirmed, so this is presented as a plan). Going further back, a March 2013 S&P AutoTechInsight article reported that Sumitomo Electric and Sumitomo Wiring Systems would set up a plant for aluminum harnesses in Thailand (scheduled to start operating in September 2014). This is old news and is mentioned only as background.

What Determines Crimp Quality, and Why You Cannot Tell from the Outside

Crimping is a method in which the terminal barrel (the tubular part) wraps around the wire’s conductor strands and is compressed by a tool to create an electrical and mechanical connection. The main factors that affect quality are as follows.

• Crimp height and crimp width: The dimensions after crimping. Too little compression gives weak contact; too much damages the strands
• Degree of strand compression: Whether the strands are packed tightly together inside the barrel with no gaps
• Insulation crimp: The part crimped over the insulation so that wire movement is not transmitted to the conductor crimp
• Bellmouth: The flared, trumpet-shaped end of the barrel. Its shape is there to avoid damaging the strands

The crimping process itself can also be viewed as a change in force. A paper (July 2025) by a research group at Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) in Germany divides the crimp force curve into 4 phases: Centring, Rolling in, Compression and Springback. A CFM is a mechanism that looks at how much the shape of this curve differs from the curve for a good crimp.

The difficulty is that many crimp defects are hard to see from the outside. Even if several strands are cut, or insulation is caught in the barrel, the crimp can look normal. According to an OES Technologies datasheet (April 2024), the detection targets of its CFM2100R include insulation caught in the conductor crimp, missing strands, changes in crimp dimensions, deformation of the crimp, terminal twisting, tool wear or breakage, changes in crimp materials, and bellmouth variation. This is a manufacturer’s announcement and does not mean that every defect can be detected under every condition. It does, however, give a sense of how many defects are hard to catch with visual inspection alone.

That is why harness plants have long continued tests that cut the crimp open to look inside, and tests that pull the wire to measure the force at which it comes out. Monitoring does not replace them; it is a separate means of watching trends across all crimps without destroying them.

Comparing Crimp Monitoring Methods

“Crimp monitoring” covers several methods, and what you can learn depends on the method. Here are 5 representative ones side by side.

A TE white paper (2005) illustrates this division of roles well. As the manufacturer’s view, it states that crimp force monitoring can check every crimp without destroying it but does not confirm whether the crimp matches the terminal’s design intent; that crimp height alone cannot distinguish process variables such as variation in the wire or terminal; and that pull testing may not be sensitive enough to detect over-compression. It presents the idea of combining crimp height and crimp force. It is an old document, but the point that each method shows different things still holds today.

Product examples are given below as manufacturers’ announcements.

• For Komax’s crimp force monitor MicroForce 85, according to the manufacturer, it minimizes defective crimps, helps optimize the process through data analysis, and ensures production traceability.
• For Komax’s C1360 crimp module, according to the manufacturer, it has a crimp force of 22 kN and an integrated crimp force monitor, and automatically adjusts crimp height with a spindle drive. Its conductor range is 0.125 to 6.0 mm².
• For Komax’s crimp height measuring device Komax 341, according to the manufacturer, it measures terminals up to 8 mm in height and compares them with the target crimp height, with a resolution of 0.001 mm and data transfer via an RS 232 interface.
• For Komax’s (Schleuniger line) CHM, according to the manufacturer, it can be combined with the measuring equipment of crimp force monitors and pull testers via WinCrimp software.
• For TE Connectivity’s CQM, according to the manufacturer, it positions crimp height measurement as the most effective non-destructive method, states that repeat measurements within 0.12 mm are possible for each analyzed crimp, and says it also supports crimping machines not made by TE.
• For Schäfer’s CrimpWise, according to an explanation contributed by a company engineer to Wiring Harness News, it continuously measures the force profile of each crimp and compares it with a taught reference curve; missing strands often lower the peak force, and caught insulation often raises displacement and peak force.
• According to a catalog (2022) for the fully automatic terminal crimping machine CPR-ZERO posted on Japan Automatic Machine’s website, it covers wire of 0.3 to 2.0 mm², and a pressure control unit (Crimp force monitor) and a crimp camera that visually inspects the crimp shape are available as options.

All of these are manufacturers’ descriptions, not results of third-party verification. The purpose of the comparison is to understand “which method tells you what,” not to rank models.

In the sense of judging each cycle as good or bad from a force waveform, crimping is the same type of process as tightening and press-fitting. For the thinking behind waveform judgment, our article on selecting servo press press-fit equipment is also a useful reference.

What to Record: Data Items per Crimp and per Setup

Whether monitoring succeeds depends half on how you design what to record. Consider the items in two groups: “required” and “recommended.”

A useful reference for thinking about record items is LV 214-4, a terminal test specification of German automakers. According to an explanatory article by Schleuniger (undated, describing the specification as a draft at the time), the items to record are the terminal ID, crimp data, wire cross-sectional area, wire type, the machine, applicator and crimp force monitor used, relative deviation, headroom, and the results of the feasibility test. Whether this matches the requirements of the current version must be confirmed against the original text and customer specifications, but the backbone of keeping together “which machine, which die and which monitoring device crimped which wire and terminal” can be used as is.

Manufacturers’ announcements include examples pointing in the same direction. TE’s 2005 white paper describes entering part number identification and batch size on screen so that operation is managed in an auditable form, and the OES Technologies CFM2100R datasheet states that selecting a part number loads the settings automatically. The further part number selection and recording move away from human handwriting, the fewer mix-ups there are.

What matters is not so much the crimp force or crimp height values themselves, but having the “crimp ID,” “setup number,” “lot” and “time” all in place. Even if the values are kept perfectly, they cannot be used in an explanation to the customer if you cannot tell which harness the wire went into.

Designing Wire Harness Traceability: Linking Crimp Data to Harness IDs

The difficulty with harnesses is that crimped wires move to the next process as bundles without each wire carrying its own serial number. For this reason, the links are built in stages, as follows.

1. At setup: Scan the terminal reel label and the wire drum label, and record them under the setup number together with the applicator ID and machine ID
2. Cutting and crimping: Attach the per-crimp data (CFM judgment, time) to that setup number
3. Wire bundle: Put a label on each bundle of wires that has finished cutting and crimping, and link the bundle ID to the setup number
4. Assembly board: Read the IDs of the bundles used on the assembly board, assign a harness ID and issue a label
5. Shipping: Link the harness ID to the shipping lot and the delivery destination

With this in place, you can trace back from the harness ID to the bundle, the setup, and the terminal reel and wire lot. Conversely, when a problem is found with a particular terminal reel, you can identify the harnesses that used it. The idea of linking equipment history to parts is covered in detail in our article on production equipment traceability (machine-to-part genealogy) and our article on connecting equipment history with production lots.

For assigning harness IDs, see our article on serial number management systems; for issuing labels, see our article on label printing systems. If you mark connector housings or protective parts directly, also check our article on selecting industrial laser markers.

Time is easy to overlook. If the clocks of the crimping machines, label printers, scanners and servers are out of sync, crimp data from around a changeover can be assigned to the wrong setup. Keeping the setup number in the machine-side records as well, so the design does not rely on time alone, makes it more robust against clock drift. Methods for synchronizing time and the related acceptance tests are summarized in our article on factory time synchronization.

Granularity also needs to be decided. Do you link every per-crimp record all the way to the harness ID, or link by setup? The former allows finer quarantine, but increases the effort of managing and reading bundles. Neither is correct across the board; decide based on customer requirements and the process risk of each circuit. Retention periods are the same: decide them based on customer requirements and your internal document control rules. If retention periods differ by customer, a design that can hold a retention period for each part number will save trouble later.

Software examples are also given as manufacturers’ announcements. For Komax’s WPCS (Wire Processing Communication Standard), according to the manufacturer’s Japanese-language page, it is a flexible interface that enables simple networking of fully automatic wire processing machines. For DiIT, the company’s MES for the harness industry, according to the manufacturer, it guarantees end-to-end traceability. The pages do not state which data can be exchanged in which format, so if you are considering adoption, you need to check with the vendor for each model and generation of your existing machines.

Setup and Applicator Management: First-Piece Checks, Hit Counts and Replacement

Crimping conditions change at every setup. Because the combination of wire, terminal and applicator changes, the setup is a quality boundary.

As one example of check frequency at setup, consider Molex’s “Quality Crimping Handbook” (1996, for users of Molex terminals and tools). In the handbook’s example, at least 5 samples are measured for crimp height in the setup check, and at least 25 points are measured for process capability evaluation. During production, crimp height is checked every 250 to 500 pieces, and a Cpk above 1.33 is said to be acceptable for most applications. The handbook itself notes that its content may differ somewhat from other companies’ procedures, so read it as one example based on Molex products, not as an industry-wide standard.

The same handbook lists wire changes, terminal lot changes, tool damage from machine jams and operator changes as examples of causes of process drift. With setup number and lot records, each of these lets you trace afterwards “when and what changed.” Changes of terminal or wire supplier and applicator replacement are treated as 4M changes, and, as in our article on 4M change management systems, the change boundary and the first-piece check are kept on record.

Handheld terminals can be used for scanning at setup. If the system reads the terminal reel and wire drum labels and stops on the spot when the part number differs from the setup instruction, mix-ups are prevented at the entrance. How to choose terminals and run acceptance tests is covered in our article on implementing handheld terminals.

Applicator hit counts are also something to record. As hits accumulate, blades and dies wear, and the trends in crimp height and crimp force gradually change. One example that illustrates the idea of managing hit counts is Weidmüller’s Crimp Counter for manual crimping tools. According to the manufacturer’s flyer (August 2025), it has a resettable service counter and a total counter that records every crimp over the tool’s entire life, the maintenance interval can be selected from 5,000, 10,000 or 15,000 cycles (default 10,000), and it can be read via NFC. This is a product for hand tools and does not provide a basis for the life of applicators on automatic machines. The idea of “holding a hit count for each tool and deciding maintenance timing from records,” however, applies directly to applicator management.

Items worth recording for applicator management are the applicator ID, the machine it is mounted on, the cumulative hit count, hits since the last maintenance, the history of maintenance and part replacement, and the results of the first-piece check after maintenance. Decide hit count limits and maintenance intervals from terminal manufacturers’ documents and your own track record, and keep the basis on record as well.

Crimp Cross-Section Analysis and Terminal Pull-Out Testing: The Role of Destructive Tests

Installing a CFM does not eliminate destructive testing. Cross-section analysis and pull testing are the means of directly confirming what is inside a crimp and how strong it is.

Cross-section analysis (micrograph) is a test in which the crimp is cut, polished and its cross-section examined under a microscope. According to a Wiring Harness News article (April 2023), the items evaluated are the degree of strand compression, the curl of the crimp flanks, damage, symmetry and voids. As a simplified method, the crimp is cut with a saw (nippers cannot be used because they deform the crimp), then coarse polishing, fine polishing and electrolytic cleaning follow in that order. The article notes, however, that a simplified cross-section only shows trends and does not produce measurements or reports with the same accuracy as laboratory analysis.

As an example of frequency, a Schleuniger representative said in another article that VW requires a cross-section at every setup, and that there were cases where the frequency of cross-section analysis rose from a few times a week to 20, 30 or 40 times a day. The representative also said that automotive standards are much stricter and require clean B-shaped symmetry. These are a representative’s remarks, not the text of a standard. Decide your own frequency according to customer requirements.

Care is needed with numerical pass/fail criteria for cross-sections. Example criteria such as void ratios in the cross-section can be found online, but some of them are based on ECSS-Q-ST-70-26C, a space standard of the ESA (European Space Agency). Space criteria cannot be used as is as pass/fail criteria for automotive harnesses. Confirm criteria against terminal manufacturers’ specifications and customer specifications.

Pull testing is a test in which the crimped wire is pulled and the force until it pulls out or breaks is measured. In the procedure in Molex’s handbook (1996), about 152 mm of wire is used, the insulation crimp is made ineffective, and the test is run at a pull speed of 50 mm/min. At least 5 wires are used for the setup check and at least 25 for process capability evaluation, and the lowest value is compared with the specification minimum. The handbook also states that pull force is only a minimum-value requirement.

The same handbook includes a reference table of pull-out forces citing UL 486A values. Some examples are as follows.

This is a reference table from the 1996 Molex handbook, and the table itself carries a note to “refer to each specification.” These are not values from current UL standards, IEC 60352-2, USCAR-21 or customer specifications. Follow each specification for pass/fail decisions.

Interpower, a power cord manufacturer, states on its quality page that it evaluates 4 items (crimp height, crimp width, compression of the conductor within the terminal crimp, and conductor pull-out force) and that it complies with terminal manufacturers’ specifications, UL 486A-486B and IEC 60352-2. Although the industry is different, the structure of combining dimensions, cross-sections and pull tests is the same.

For micrometers, crimp height gauges, pull testers and CFMs alike, measurement reliability is the foundation for explaining quality. How to manage calibration due dates and results through records is summarized in our article on calibration management systems.

Standards and Customer Requirements: IEC, IPC/WHMA, USCAR, LV 214 and JIS

The main standards related to crimping are organized below by number, title and scope. Confirm the details of requirements and numerical pass/fail values against the original texts and customer specifications.

On USCAR-21, SAE’s description states that it is a test that simulates in the laboratory the stresses a crimp experiences over a typical life (15 years, 150,000 miles). It also states that it has tests capable of detecting deficiencies in crimp tool geometry, plating quality, strand distribution and the strand count of the cable.

JIS C 2805 deserves attention. Judging from the name alone, it looks like a standard for crimp terminals in general, but it covers bare crimp terminals and insulated crimp terminals used for wiring electric power equipment. If your internal documents cite it as the acceptance criterion for automotive harness terminals, it would be advisable to review that.

The scope of IEC 60352-2 is important in the same sense. It covers stranded wire only up to 10 mm², so EV high-voltage large-gauge wire is outside its scope. Evaluation of large-gauge wire needs to be decided separately based on terminal manufacturers’ and customers’ specifications.

How to demonstrate in audits the records required by standards and customer specifications is covered in our article on IATF 16949 traceability. Linking crimp data becomes part of that evidence trail.

CFM Limitations and Handling False Alarms

A CFM is a powerful tool for watching every crimp without destroying it, but it has limitations. In its related-work section, the FAU paper introduced earlier points out that commercial CFMs rely on manually set tolerance limits; that, as a result, slight process variations trigger false alarms; that, conversely, they cannot detect actual defects that produce only slight deviations; and that frequent readjustment is required.

Put in shop floor terms, if you narrow the limits, false alarms stop the machine and operators end up widening the limits. If you widen them, defects slip through. It is a tug of war. The basic countermeasures are these 3:

• For each terminal, first confirm whether the CFM can detect defects: This corresponds to the thinking behind LV 214-4. According to an explanatory article by Schleuniger (undated, describing the specification as a draft at the time), with the smallest permissible wire for the target terminal, it confirms that the relative deviation of peak force is within 1% over 300 good crimps, that 9% missing strands (1 strand for a 7-strand wire, 3 strands for a 32-strand wire) is detected in all 10 samples, that 30% caught insulation is detected in all 10 wires, and that headroom exceeds 35%. These are figures as of the explanatory article and cannot be asserted to be current requirements, but the idea of “deliberately making defects with your own terminal and wire combinations and testing whether they are detected” is usable
• Keep records of limit changes: Record who changed the limits, when, and in which setup, and configure the system so that changes require approval
• Decide the re-teaching procedure: Decide in advance whether to re-teach the reference when the wire lot or terminal lot changes, and how many crimps to use for teaching

Research on using AI to reduce false alarms and missed detections is also progressing. The FAU paper (July 2025) classifies good crimps, missing strands and caught insulation from crimp force curves, and reports a fault detection accuracy of 95.9%. The scope, however, is limited. The subjects were 2 types of copper wire (0.50 mm² with 16 strands and 0.35 mm² with 12 strands), with a total of 1,617 crimp force curves measured on a single semi-automatic crimping press fitted with a force sensor. The defects were made intentionally. In testing, 6 of 161 good crimps were judged defective, and 4 of 163 defective crimps were judged good (these 4 are described as having been made as minor defects at a level that current CFMs also cannot detect). It cannot be said that the same accuracy would be achieved as is on production lines.

As prior work, the same paper cites values such as 86.9 to 100% for multi-class classification of fastening, press-fitting and crimping using a 1D CNN, and up to 99.7% for a combination of CFM and multi-camera optical inspection. These are secondary citations from this paper, and each value is based on that study’s own data, so they cannot be compared side by side.

For watching crimp height values on control charts and catching signs of drift early, our article on SPC anomaly detection AI is a useful reference. Before AI, the first step is to reliably collect CFM judgments and crimp height values and get them into a state where they can be viewed on control charts.

Responding to Abnormalities: Stop, Quarantine, Cut

When a CFM gives an NG, whether the machine automatically cuts or ejects that wire, and how it does so, differs by model. Check this for each model when you build the list of existing machines. On bench presses, the procedure is for the operator to cut the NG wire on the spot and put it in the NG bin. If there is no record of whether the wire was actually cut, you cannot explain it later.

When abnormalities occur, the most effective thing is the lot and setup records. For example, when a defect is found in a particular terminal reel, records let you narrow the scope in the following order.

1. From the lot of the terminal reel where the defect occurred, look up the setup numbers that used it
2. From the setup numbers, look up the wire bundles crimped in those setups
3. From the bundle IDs, look up the harness IDs and shipping lots that used them

Without records, you have no choice but to suspect every harness made during that period. The design of forward and backward tracing is covered in our article on trace-forward / trace-back, and procedures for responding to customers and recalls in our article on recall traceability.

Records are also needed when defective crimps are cut out and reworked. Which circuit of which harness was redone, and which inspection was carried out after rework? Without this on record, reworked products cannot be distinguished from regular products. For the flow from quarantine to re-inspection, see our article on rework management systems.

EV High-Voltage, Large-Gauge and Aluminum Wire

As EVs spread, the wire that harness plants handle is changing. A TE page states that emissions regulations have led to rapid adoption of EVs, HEVs and PHEVs, which in turn has led to a surge in the use of large-gauge wire and cable. As challenges, it lists clocking, crimp force, extrusion, cracking, crimp height, cross-section analysis, water ingress, and the differences between aluminum and copper conductors, among others.

A TE Connectivity engineer said in a Wiring Harness News article (undated) that low-voltage wire is usually crimped at under 3,000 lbs, whereas mid-size high-voltage wire of 6 AWG and larger requires about 10,000 lbs, and 90 to 120 mm² requires up to 40,000 lbs. The engineer also mentioned that 70 mm² wire is extruded by about 5 mm, that the thicker the wire, the slower the optimal crimping speed, and that orientation (clocking) becomes an issue with asymmetric terminals, and stated that a 2 m length of 50 mm² cable weighs 1.5 kg. On that basis, the engineer recommends crimp quality monitoring and camera checks before and after crimping. These are remarks by a manufacturer’s representative.

Aluminum wire brings yet another set of challenges. An SAE paper (2012) by researchers from AutoNetworks Technologies and Sumitomo Wiring Systems states that weight reduction and rising copper prices motivated the development of aluminum harnesses, and that because of the oxide film on aluminum, terminal connection was a challenge and aluminum had conventionally been limited to battery cables; the researchers analyzed the mechanism of crimp connection and developed dedicated terminals. The abstract gives no numerical values.

For large-gauge aluminum wire, ultrasonic welding is also available as a method separate from crimping. Schunk Sonosystems’ use-case page presents an example of welding aluminum cable to copper alloy terminals with cross-sections up to 85 mm². This is a manufacturer’s use-case presentation, not a comparison of its merits against crimping.

With large-gauge and aluminum wire, the point to watch is the scope of the standards. IEC 60352-2 covers stranded wire up to 10 mm², and USCAR-21 assumes automotive stranded copper wire at low voltage (0 to 48 VDC). Large-gauge high-voltage wire falls outside both scopes. Aluminum wire is also not the stranded copper wire that USCAR-21 assumes. Crimp parameters and CFM limits will need to be set starting from the terminal manufacturer’s recommended conditions, while recording the results of cross-section analysis and pull testing. Keeping those conditions and their basis for each terminal and wire combination provides the material for explaining things to the customer later.

Requirements to Write into the RFP

When ordering a system for collecting and linking crimp data, write the following items into the RFP (request for proposal) and confirm them through the vendors’ responses.

For the thinking behind a platform that collects actuals from machines, see our article on automatic collection of production actuals; if you keep time-series data such as waveforms over the long term, our article on introducing an industrial historian is also a useful reference.

The most important thing in the RFP is for the ordering side to prepare the list of target machines. If you send it out after listing model, year of manufacture, firmware, communication ports and CFM presence, the vendors’ responses come back at a consistent level of detail and are easier to compare.

FAT/SAT and Acceptance Criteria

Consider acceptance testing in two parts: the FAT (factory acceptance test), performed before shipment, and the SAT (site acceptance test) performed on the line at your site. Acceptance target values are not set as an industry standard, so decide them yourself and put example targets in the RFP.

FAT (verification with simulated data)

• Feed in simulated crimp data and confirm that zero records are missing
• Confirm that the same judgment is reproduced when the same waveform or measured value is input
• Confirm that crimping can be stopped when the part number scanned at setup is different
• Confirm that limit changes are kept in the history and cannot be made without approval
• Confirm that reports can be output in the customer’s format

SAT (verification on the actual line)

• Over a set period of operation, the crimp record rate and lot linking rate meet your own targets
• Clock drift between machines, scanners and servers stays within a range that does not cause setups to be misassigned
• Alarms for CFM NGs and missing records reach the designated recipients
• Results of existing cross-section analysis and pull tests can be matched against the crimp data from the same setup
• With representative terminal and wire combinations, deliberately create missing strands and caught insulation and confirm whether the CFM can detect them (applying the feasibility test idea from the LV 214-4 explanatory article to your own conditions)

The last item is especially important. Simply having a CFM installed means little; only once you have confirmed that it can catch defects with your own terminals and wires can it be used to explain quality.

Cost and Benefit Model Calculation (Assumed Numbers)

All figures from here on are assumed numbers prepared for this article. Official prices for CFMs, gauges and MES could not be confirmed, so costs are also placeholders. Use this as a framework when doing your own calculation.

Assumed plant: A Japanese-affiliated automotive wire harness plant in Thailand. 30 fully automatic cut-and-crimp machines (with built-in CFM) and 20 bench presses (without CFM), 2 shifts, 250 operating days per year.

Current costs (baseline amounts, annual)

Configuration A: Data Collection from the 30 Existing Fully Automatic Machines and Linking to Lots and Harness IDs

Install software that collects CFM and measurement data from the 30 existing fully automatic machines and links it to lots and harness IDs, plus scanning of terminal and wire lots. No new crimping machine hardware is purchased. Assume an initial cost of 2,400,000 THB and an annual operating cost of 300,000 THB.

The net effect is 990,000 − 300,000 = 690,000 THB/year, and the payback period is 2,400,000 ÷ 690,000 ≈ about 3.5 years.

Configuration B: Configuration A Plus CFM Retrofits on the 20 Bench Presses

This is evaluated as an addition after Configuration A has been installed. Assume a CFM retrofit costs 150,000 THB per unit, so the additional initial cost is 20 × 150,000 = 3,000,000 THB, and the additional annual operating cost is 100,000 THB.

• A further 15% reduction in internal rework and scrap against the baseline of 1,800,000 THB: 270,000 THB
• A further 10% reduction in claim handling against the baseline of 1,200,000 THB: 120,000 THB
• Total additional effect: 270,000 + 120,000 = 390,000 THB/year

The net effect of the addition is 390,000 − 100,000 = 290,000 THB/year, and the payback period for the addition is 3,000,000 ÷ 290,000 ≈ about 10.3 years.

Sensitivity: If the Claim Baseline Were Twice as Large (Configuration A)

If the baseline for customer claim handling were twice as large, at 2,400,000 THB/year, the 30% claim reduction in Configuration A would become 720,000 THB. The total effect would be 360,000 + 720,000 + 270,000 = 1,350,000 THB/year, the net effect 1,350,000 − 300,000 = 1,050,000 THB/year, and the payback period 2,400,000 ÷ 1,050,000 ≈ about 2.3 years.

What the Calculation Shows

• The core of the effect is collecting data from existing fully automatic machines and linking it to lots and harness IDs (Configuration A). Without buying new hardware, it affects both record-keeping effort and the scope of claim handling
• Retrofitting CFMs uniformly to all bench presses takes about 10.3 years to pay back the addition under these conditions. A decision is needed to limit it to presses handling critical circuits such as airbag, brake and high-voltage circuits
• The larger a plant’s claim handling costs, the larger the effect of linking. In the sensitivity case, the payback period shrinks to about 2.3 years

The addition in Configuration B is set as a percentage of the baseline, not of the amount remaining after Configuration A. Opportunity losses from line stoppages, winning new orders and reductions in audit workload are not included in this calculation.

Frequently Asked Questions

Can crimp monitoring alone guarantee quality?

No. A CFM is a means of looking at the crimp force curve of every crimp and picking up deviations from the good-crimp reference, but it does not confirm whether the crimp matches the terminal’s design intent. Research has also pointed out that it can miss defects that produce only slight deviations. Only by combining it with crimp height measurement, cross-section analysis and pull testing, and further linking crimp data to lots and harness IDs, can you explain quality to the customer.

What defects does a crimp force monitor (CFM) find?

Manufacturers’ announcements list missing strands, caught insulation, changes in crimp dimensions, and tool wear or breakage among the detection targets. However, whether a defect can be detected depends on the terminal and wire combination and on how the limits are set. We recommend deliberately creating defects with your own combinations and confirming whether they can be detected.

How often should crimp height be measured?

There is no single correct answer. Follow the terminal manufacturer’s specifications and customer requirements. As one example, Molex’s 1996 handbook gives at least 5 samples for the setup check and checks every 250 to 500 pieces during production. This is an example based on Molex products. Once you have decided the frequency, a mechanism that keeps the measured values linked to setup numbers rather than handwritten makes them data you can use later.

Is crimp cross-section analysis necessary?

It is necessary even with a CFM installed. Cross-section analysis is one of the few means of directly seeing what is inside a crimp, such as strand compression, symmetry and voids, and it also provides material for confirming that the CFM reference is correct. The frequency is determined by customer requirements. There have also been reports of automakers requiring a cross-section at every setup. Confirm numerical pass/fail criteria against terminal manufacturers’ and customers’ specifications, and take care not to borrow criteria from space standards.

What are the criteria for terminal pull-out testing?

Follow the terminal manufacturer’s specifications and customer specifications. Pull-out force tables found online include old reference values, such as the UL 486A values cited in Molex’s 1996 handbook. Do not use them as pass/fail criteria as they are; check their source and date. Keeping test conditions consistent (wire length, treatment of the insulation crimp, pull speed) is also important for comparing results.

How far does wire harness traceability need to go?

It depends on customer requirements and how narrowly you want to be able to limit the quarantine scope when something goes wrong. Linking every per-crimp record all the way to the harness ID is the finest level, and linking by setup is the option one step short of that. At a minimum, being able to trace between the terminal reel lot, wire lot, applicator ID, machine ID, setup number and harness ID lets you narrow the scope when a problem occurs. Retention periods are decided by customer requirements and internal rules.

Do bench presses need CFMs too?

Not necessarily on every unit. In this article’s calculation with assumed numbers, a uniform retrofit to 20 bench presses took about 10.3 years to pay back the addition. A realistic approach is to first bring setup scanning, crimp height records and operator records to bench presses as well, and to consider CFMs only for presses handling critical circuits such as airbag, brake and high-voltage circuits. If the customer requires 100% monitoring for certain circuits, those come first.

Summary

• Crimp quality control is not “done once you install a CFM.” CFM, crimp height measurement, cross-section analysis and pull testing show different things, and none of them replaces another
• What to decide first are 3 things: the circuits and terminals under focused control; the record items per crimp (or per setup) and how they are linked to lots and harness IDs; and the records of applicator hit counts and first-piece checks
• Judging from manufacturers’ announcements, many fully automatic machines can handle CFM and measurement data. First collect data from existing machines, and link terminal reels and wire lots through scanning at setup
• Build the links by connecting setup numbers and IDs at each stage: setup, wire bundle, assembly board and shipping. Do not rely on time alone
• CFMs have limitations: false alarms, missed detections and the effort of readjustment. Deliberately create defects with your own terminals and wires and confirm whether they can be detected
• Grasp the overall picture of standards by number and scope. JIS C 2805 is not a standard for automotive open-barrel terminals, and IEC 60352-2 covers up to 10 mm², so large-gauge wire is outside its scope
• In the calculation with assumed numbers, the core of the effect was data collection from existing machines and linking (Configuration A). Consider CFM retrofits on bench presses only for critical circuits

TOMAS TECH can help even at an early stage, such as taking stock of which circuits should receive focused control and which data can be obtained from your existing crimping machines, or organizing RFP requirements. Please contact us through this form.

References

• https://webstore.iec.ch/en/publication/68899
• https://www.evs.ee/en/iec-60352-2-2024
• https://www.dinmedia.de/en/standard/ipc-whma-a-620f/398022844
• https://www.technickenormy.cz/en/ipc-whma-a-620-revision-f/
• https://saemobilus.sae.org/standards/uscar21-4-performance-specification-cable-terminal-electrical-crimps
• https://saemobilus.sae.org/content/uscar2-8
• https://webdesk.jsa.or.jp/preview/pre_jis_c_02805_000_000_2010_j_ed10_ch.pdf
• https://www.circuitnet.com/news/uploads/3/Article_LV_214_Final.pdf
• https://www.komaxgroup.com/products/quality-tools/microforce-85
• https://www.komaxgroup.com/products/wire-processing/higher-automation-platforms/process-modules/c1360
• https://komaxgroup.com/products/quality-tools/komax-341
• https://www.komaxgroup.com/products/testing-and-quality-tools/cable-and-harness-quality-tools/crimp-height-measuring/chm
• https://komaxgroup.com/ja-jp/products/software-and-networking/wpcs
• https://www.komaxgroup.com/brands/diit
• https://www.te.com/en/products/application-tooling/intersection/crimp-quality-monitoring-cqm.html
• https://www.te.com.cn/content/dam/te-com/documents/automotive/global/at-wp-crimp-height-meeting-contemporary-quality-standards.pdf
• https://www.te.com/en/campaigns/transportation-solutions/gatd-high-voltage-wire-processing-equipment.html
• https://wiringharnessnews.com/monitoring-crimp-quality-in-real-time/
• https://www.oes-inc.com/app/uploads/sites/3/2024/04/CFM2100R-April-2024-EN.pdf
• https://jam-net.co.jp/wp-content/uploads/2024/08/CPR-ZERO_catalog.pdf
• https://assets.dam.weidmueller.com/assets/api/8c623cf0-e820-4f66-9c07-bef9a7b46b48/Original/small_Flyer_CrimpCounter_EN_WEB.pdf
• https://schunk-group.com/sonosystems/en/use-cases/large-cable-cross-section-for-energy-supply-in-the-vehicle-electrical-system~u11708
• https://www.marklines.com/en/news/351560
• https://www.kaohooninternational.com/?p=592011
• https://www.boi.go.th/upload/content/PR153_2569EN.pdf
• https://www.boi.go.th/upload/content/PR101_2569EN.pdf
• https://www.krungsri.com/en/research/industry/industry-outlook/hi-tech-industries/auto-parts/io/auto-parts-2026
• https://www.nationthailand.com/business/automobile/40071777
• https://www.nationthailand.com/business/40071775
• https://www.nationthailand.com/business/automobile/40071986
• https://finance.yahoo.com/economy/policy/articles/thai-auto-sector-facing-crisis-082852558.html
• https://www.alcircle.com/news/yazaki-expands-its-aluminium-wire-production-capacity-103073
• https://autotechinsight.spglobal.com/news/20525/sumitomo-electric-and-sumitomo-wiring-team-up-to-establish-aluminium-harnesses-plant-in-thailand
• https://arxiv.org/pdf/2507.00102
• https://saemobilus.sae.org/papers/crimping-technology-aluminum-wire-automotive-wire-harness-2012-01-0922
• https://wiringharnessnews.com/intricacies-of-high-voltage-cable-processing/
• https://cdn.shopify.com/s/files/1/0838/3732/1530/files/qual_crimp.pdf
• https://www.interpower.com/ic/designers/QualityTopics/Crimp-Cross-Section-and-Pull-Test.html
• https://www.aac-research.at/microanalytical-assessment-of-crimp-joints/
• https://wiringharnessnews.com/micrograph-the-inner-values-of-a-crimp
• https://wiringharnessnews.com/developments-in-cross-section-analysis/