Cable Harness Overmold and Potting: Adhesion, Cure, Voids and Strain-Relief RFQ Evidence

A smooth overmold can hide a poor bond, an off-ratio compound, incomplete cure, trapped air or a cable load carried by the conductor terminations instead of the jacket. A quotation that says “waterproof molded connector” does not define the material interfaces, process window, internal fill or mechanical load path.

For procurement, the direct answer is: freeze the complete assembly and material stack, then require evidence for surface preparation, metering and mixing, dispense or molding conditions, cure release, void acceptance and the finished cable’s axial, torsional and bending behavior. Review the evidence after relevant environmental conditioning, not only on a freshly molded cosmetic sample. Keep sealing, adhesion, electrical insulation and strain relief as related but separately demonstrated functions.

This guide focuses on polymer applied around a cable termination, splice, backshell, connector rear or small harness transition. It complements the cable strain-relief and bend-radius guide, the dynamic-flex harness guide, the terminal-retention guide and the sealed-connector leak-test guide. Those articles address route, motion, terminal locking and leak-test correlation. Here the purchasing decision is whether the molded or potted transition is reproducible and supported by evidence for its declared functions.

No part of this article confirms a SINAWATTS material, molding process, potting process, test capability, certification, ingress rating, inventory, price, MOQ, lead time or customer result. Material and process suitability must be established for the exact cable, housing, compound, geometry, environment and production site.

Direct answer: what should the RFQ define?

A useful RFQ should require one controlled package containing:

• assembly drawing, section views and dimensions of the finished overmold or potting cavity;
• connector, terminal, splice, wire and cable part numbers and revisions;
• conductor construction, insulation and jacket polymer, color, nominal dimensions and tolerances;
• housing, insert, backshell, pre-mold, grommet and seal materials where they affect compatibility;
• exact overmold, encapsulant, potting compound, hardener, primer and cleaning-material order codes;
• approved material storage, conditioning, shelf-life and thaw rules;
• surface-preparation method and the maximum permitted delay before molding or dispensing;
• resin/hardener ratio, metering tolerance, mixing method, purge rule and batch traceability for two-part materials;
• material, component, tool and ambient temperatures;
• injection or dispense pressure, speed, shot/dispense quantity, vacuum or venting method and relevant timers;
• fixture, cable support and conductor position during fill and cure;
• cure schedule and measurable production-release criterion;
• adhesion, fill, void, flash, crack and interface acceptance criteria;
• axial pull, torsion, bend/flex and connector/terminal post-test checks where required;
• initial and post-conditioning evidence on serialized samples;
• production inspection frequency, reaction plan and data retention; and
• change-notification and revalidation rules.

The buyer should be able to identify what bonds chemically, what is retained mechanically, what is merely enclosed and where an external cable load travels. If the section drawing and load path are missing, a glossy exterior cannot close the technical review.

Distinguish overmolding, potting, encapsulation and strain relief

Suppliers may use these terms differently, so define them on the drawing.

For this RFQ framework:

• overmolding means forming polymer around an existing cable/connector subassembly in a mold, commonly by an injection or low-pressure molding process;
• potting means dispensing a flowable polymer system into a cavity where it cures;
• encapsulation means enclosing the relevant component or assembly in polymer to the defined extent;
• adhesion means resistance to separation at a named material interface under a stated method and condition;
• mechanical interlock means retention created by undercuts, holes, ribs, knurls or other geometry even when chemical adhesion is limited;
• strain relief means transferring external load away from the conductor or terminal interface through a controlled cable-to-structure load path; and
• sealing means limiting ingress or leakage at a defined boundary under a defined test, not simply covering a surface.

NASA-STD-8739.1B with Change 2 defines encapsulation as complete encasement in resin and notes that “potting,” “embedment” and “molding” are other terms used in the electronics industry. NASA’s official active-standard record, checked on 2026-10-04, states that the document covers polymeric applications for electronic assemblies. That NASA definition is useful for vocabulary and evidence discipline; it is not a universal commercial-harness specification and should not be imposed unless the contract invokes it.

IPC’s current IPC/WHMA-A-620F table of contents, dated October 2025 and checked on 2026-10-04, identifies the document as Requirements and Acceptance for Cable and Wire Harness Assemblies. It includes sections for potting (thermoset molding), fill, fit to wire or cable, cure and over-molding, while its published scope says the standard does not provide cross-section or X-ray acceptance criteria. IPC also explains publicly that A-620 does not itself set every in-process or product-inspection frequency. Therefore an RFQ should state the required revision/class where applicable and still define sampling, product-specific criteria and records.

Start with function, not a preferred compound family

Write what the polymer must do in the actual application. Possible functions include:

• reduce bending stress at the connector/cable transition;
• retain the cable jacket against axial or torsional load;
• protect splices or terminals from handling;
• maintain conductor spacing;
• exclude water, dust or process fluid at a defined boundary;
• resist vibration, shock or repeated flexing;
• provide electrical insulation;
• distribute heat or avoid damaging a heat-sensitive component;
• resist named chemicals, UV or temperature cycles;
• support a required color, marking or keying feature; and
• enable repeatable assembly without contaminating mating or sealing surfaces.

Do not assume one compound delivers every function. A hard epoxy may provide a rigid encapsulation but transfer bending to a sharp exit. A soft elastomer may reduce stress yet provide different retention, chemical or abrasion behavior. A thermoplastic overmold may grip through geometry while showing limited chemical adhesion to a low-surface-energy cable jacket. A sealed appearance can coexist with a leak path along stranded conductors or an unbonded interface.

Assign an evidence type to each declared function. For example, an axial retention claim needs a force/displacement method and post-test checks. An ingress claim needs a controlled boundary and correlated test. An insulation claim needs the applicable electrical test. A cosmetic inspection cannot serve all three.

Molex’s official custom cable assembly page, checked on 2026-10-04, describes its custom overmolded assemblies as providing strain relief at the interface between the connector and jacketed cable and says that overmolded strain relief reduces mechanical stress on wires and terminals. This is a manufacturer statement about that offering and a useful description of intended function. It does not prove retention, sealing or cycle life for an unnamed custom assembly.

Freeze every material interface

An overmold does not bond to “a cable.” It encounters specific materials and surface histories. List each interface in the drawing and BOM:

• overmold to cable jacket;
• overmold to connector housing or backshell;
• overmold to an inner pre-mold;
• potting to housing wall;
• potting to wire insulation or cable jacket;
• potting around terminals, splices, shields or drain wires;
• primer to substrate and polymer to primer;
• polymer around a seal, grommet or insert; and
• overmold-to-overmold interface when there are multiple shots.

For every substrate, request the manufacturer, order code, polymer designation where controlled, color/additive variant, mold cavity or production source where relevant, and surface treatment. A nominal material family such as PVC, PUR, TPE, silicone, PA or PBT does not prove that two commercial formulations bond alike. Plasticizer, flame retardant, release agent, lubricant, pigment and ageing can change an interface.

The process record should control contamination from cable drawing or extrusion, connector molding, handling, tape, marker ink, cleaning solvent and mold release. If the supplier relies primarily on mechanical interlock, the section drawing should identify undercuts, holes, ribs or captured features. If adhesion is required, the validation should use the actual production substrates and preparation.

3M’s current Scotch-Weld DP270 Clear technical data sheet, revision dated May 2026 and checked on 2026-10-04, describes that named product as a two-part low-viscosity epoxy for potting, sealing and encapsulation. Its instructions state that surface-preparation effort depends on required bond strength and environmental ageing resistance and that contaminants such as oils, dust and mold-release agents must be removed for high-strength bonds. This is product-specific guidance, not a recommendation to use DP270 on a particular harness or permission to copy its preparation to another material.

Convert the material data sheet into a controlled process

A technical data sheet provides boundaries and typical properties. Production still needs a released work instruction linked to the exact material revision and package form.

For two-part potting or encapsulation, control:

• component A and B identities, lots, expiration dates and storage histories;
• preconditioning, thaw and remix requirements;
• ratio basis: mass or volume;
• ratio tolerance and how the metering system demonstrates it;
• density correction if mass and volume records are compared;
• static/dynamic mixer identity and discard/purge quantity;
• mixing time, speed and vessel geometry for hand or batch mix;
• mixed-material temperature and viscosity window;
• pot life or working time, including the defined start event;
• maximum batch size and exotherm considerations;
• vacuum degassing or another de-air method where required;
• dispense path, needle/nozzle, fill direction and interruption limits;
• fixture and cable position until gel or cure prevents movement; and
• cleanup and prevention of material entering mating or contact surfaces.

For a molded thermoplastic or reactive overmold, control the applicable resin drying/conditioning, lot, barrel/material temperature, mold temperature, injection profile, pressure, hold/cool time, vent condition, shot size, insert preheat, insert position and cavity identity. Do not invent generic molding values in the purchase drawing; require the supplier’s validated window and evidence that it produces the approved assembly.

NASA-STD-8739.1B’s official Change 2 PDF, checked on 2026-10-04, provides a rigorous example of mix-record content: unique batch identity, material identities and traceability, ratio, ambient temperature/humidity, mix time, pot life, date, operator and procedure revision. Those requirements apply within NASA’s stated scope. A commercial buyer can use the categories as an evidence model without claiming NASA compliance.

Illustrative mix-ratio check only

Assume a fictional two-part system has an approved mass ratio of 100 parts A to 20 parts B. This is not a value for any product cited in this guide.

For a planned 600 g batch:

• total ratio parts = 100 + 20 = 120;
• required A = 600 × 100/120 = 500 g; and
• required B = 600 × 20/120 = 100 g.

Suppose the batch record instead shows 510 g A and 90 g B. The actual B per 100 parts A is:

90 / 510 × 100 = 17.65 parts B per 100 parts A

Relative to the fictional 20-part target, B is about 11.8% low:

(17.65 - 20) / 20 × 100 ≈ -11.8%

That arithmetic flags a deviation; it does not determine whether the batch is usable. The material manufacturer’s allowed ratio tolerance, measuring uncertainty and approved disposition control. A supervisor should not “correct” a partly cured batch by adding hardener after the fact unless a validated instruction explicitly permits it.

Define cure by a release criterion, not elapsed time alone

Cure depends on material chemistry, ratio, temperature, mass, geometry, humidity where relevant and time. The exposed surface may feel firm while hidden material remains below its required state. Conversely, a material may continue developing properties after it is safe to demold.

Distinguish these milestones:

• end of dispense or mold fill;
• working-time or pot-life limit;
• gel or no-flow condition;
• demold or fixture-release condition;
• handling strength;
• inspection state;
• electrical-test release;
• mechanical-test release; and
• full or specified property development.

The 3M DP270 data sheet is a useful named example because it distinguishes work life and cure behavior and warns that its technical data are representative rather than specification values. A buyer should request the exact current TDS and supplier process for the offered material instead of copying a headline cure time.

Dow’s official DOWSIL TC-6020 product page, checked on 2026-10-04, identifies that named material as a two-part encapsulant, publishes a 1:1 weight mixing ratio, and lists product-specific heat-cure conditions and typical properties. These statements apply only to that product. They illustrate why the offered material identity, ratio and cure schedule must be frozen and verified from current manufacturer information; they do not define another compound’s process or acceptance limits.

Useful cure-release evidence can include an approved witness sample from the same mix, hardness or another correlated property, equipment-recorded time/temperature history, validated demold behavior and a defined inspection. The method must be shown to correlate with the hidden critical region. A surface fingerprint test is weak evidence when the critical bond is deep inside a connector backshell.

Treat adhesion as an interface and failure-mode question

“Good adhesion” is not a measurable acceptance criterion. State:

• which two materials form the interface;
• their production lots and surface histories;
• preparation and treatment-to-mold/dispense time;
• bond geometry and thickness;
• test direction, rate and fixture;
• temperature and moisture condition;
• sample age/cure state;
• displacement or strength requirement from the approved design;
• allowed failure modes; and
• post-test electrical, sealing and visual checks.

A test that tears the cable jacket may show that the interface exceeded the jacket’s local strength in that configuration. It does not produce a universal adhesion value, and it may damage the assembly before the actual service load is reached. A cohesive polymer failure, adhesive interface failure, substrate tear, geometric pullout and terminal failure have different implications. Record the failure mode and location, not just the peak force.

Use actual molded assemblies where the geometry drives the result. Flat witness coupons can help monitor surface preparation or material batches, but a coupon does not automatically represent a curved cable, a connector rib, a thick potting mass or trapped residual stress. Correlate any coupon test with the production interface it is meant to control.

Ageing matters. If the application includes thermal cycling, heat, fluid, humidity, water, salt, UV or vibration, define the applicable conditioning sequence and test the same identified specimens afterward where feasible. Keep the sequence visible: cleaning, remolding or drying a conditioned sample before adhesion assessment can erase the failure mechanism.

Define void acceptance and the inspection’s detection boundary

Air can enter during mixing, be trapped behind wires and ribs, expand during heating or remain when material cannot wet a narrow region. A void can be harmless in one cosmetic zone and critical when it creates a continuous path, exposes a conductor, reduces a bond area or concentrates stress.

Create a zone map:

• electrical-spacing zone: material is intended to maintain insulation between conductive parts;
• sealing zone: a continuous path could connect the environment to a protected cavity;
• adhesion zone: loss of bonded area affects retention or sealing;
• strain-relief zone: voids alter the intended load-transfer section;
• thermal zone: voids change a heat-transfer path where that function is required; and
• cosmetic zone: appearance matters but the void is outside the declared critical functions.

Define the allowed void size, distribution and continuous-path rule by zone from the governing design/standard. Do not apply one percentage to the complete overmold without an engineering basis.

NASA-STD-8739.1B provides a clear, bounded example for NASA electronic encapsulation. Its encapsulation section requires vacuum degassing before application and defines pre- and post-cure bubble/pinhole criteria, complete fill and intended adhesion, including a prohibition on a continuous bubble path between non-common conductors. These values and requirements belong to that NASA standard and any contract that invokes it. They are not universal acceptance numbers for automotive, marine or industrial harnesses.

Match inspection to the defect:

• external visual inspection can find surface bubbles, short shots, sink, cracks, flash and exposed areas;
• shot weight or dispense mass can indicate missing material but cannot locate a hidden void;
• sectioning can reveal a local cross-section but destroys the sample and may miss another plane;
• X-ray or computed tomography may reveal density differences, subject to material contrast, geometry, resolution and validated interpretation;
• ultrasonic or other nondestructive methods may be suitable only after method development for the specific materials;
• leak testing can reveal a connected flow path but does not map every isolated void; and
• electrical tests can detect some functional effects without proving complete fill or adhesion.

Request a detection study or reference defects near the acceptance boundary when a hidden-void method is critical. A machine image marked “PASS” is incomplete without resolution, coverage, algorithm/reviewer rules and calibration or verification evidence.

Design the fill path so evidence is repeatable

The drawing should show where material enters, how air leaves and which orientation is maintained during fill and cure. Blind pockets, closely packed wires, backshell ribs and sharp changes in section can trap air. A long pause can create a knit line or partially cured interface. Excess pressure can move wires, displace seals or force compound into a mating area.

Request these production parameters where applicable:

• fill point and sequence;
• vent or overflow location;
• component orientation;
• bottom-up or other validated fill path;
• vacuum level and time, if used;
• nozzle depth, withdrawal method and speed;
• maximum interruption between pours or shots;
• shot/dispense mass and tolerance;
• mold-clamp and insert-holding method;
• cable straightness and conductor spacing during fill;
• overflow/flash containment; and
• cleanup limits that do not damage or contaminate the interface.

3M’s current DP270 instructions require its two components to flow evenly before cartridge mixing, or to be mixed thoroughly to a uniform color for bulk or hand mixing, and they define product-specific working and cure conditions. Dow’s current TC-6020 product page separately publishes a product-specific 1:1 weight ratio and several heat-cure conditions. Use the current instructions for the selected material and equipment rather than combining isolated advice into a homemade process.

Make strain relief a visible load path

Draw the path from an external cable load into the connector or supporting structure. Label which surface grips the jacket, where the molded geometry reacts axial load and where bending begins. Show the first cable support in the installation.

A useful section should answer:

• Does the polymer grip the outer jacket or only individual insulated wires?
• Is the jacket continuous into the overmold?
• Are conductors splayed, looped or straight?
• Does a crimp, solder joint or terminal-retention feature carry load before the overmold engages?
• Is a mechanical undercut present?
• Is a flexible boot section intended to bend?
• Where does the boot transition to a rigid body?
• Can torsion rotate the cable inside the mold?
• Does the connector housing take the reaction load without distorting its seal or latch?
• Does the proposed installation support the cable before routine movement reaches the termination?

IEC’s official record for IEC 60512-17-3:2010, checked on 2026-10-04, describes a connector cable-clamp tensile test intended to assess resistance to longitudinal movement. IEC 60512-17-4:2010 describes a cable-clamp torsion-resistance test. Each applies when required by the detail specification and leaves severity and limits to the applicable specification. They provide useful test vocabulary; they do not supply a universal pull force or prove that a particular overmold qualifies as a cable clamp.

Keep contact/terminal retention separate. If an axial pull is reacted by a terminal lance before the overmold grips the jacket, the result may test the wrong interface. Compare the assembly before and after the mechanical test for conductor movement, terminal back-out, contact resistance/continuity where required, housing cracks, jacket damage, interface separation and seal performance.

Illustrative bending-moment comparison only

Assume a fictional cable experiences a 30 N lateral handling force. If the force acts 50 mm from the point where a rigid overmold ends, the simple moment at that transition is:

M = F × L = 30 N × 0.050 m = 1.5 N·m

If a controlled external support reduces the lever arm to 10 mm, the simple moment becomes:

30 N × 0.010 m = 0.3 N·m

The arithmetic shows why the first support location can matter. It does not predict stress in a real boot, approve a clamp position or provide an allowable load. Cable stiffness, boot geometry, direction, dynamic effects and material behavior must be included in the actual design and validation.

Define mechanical qualification around actual service modes

Separate these tests and decisions:

• axial pull along the cable centerline;
• lateral load at a specified distance and direction;
• cable torsion about its axis;
• repeated bend or flex about a controlled radius;
• connector mating/unmating handling;
• vibration/shock with the installed cable support; and
• abusive handling that is prohibited rather than qualified.

For each applicable test, state fixture, free cable length, connector state, load direction, rate, target, hold, cycles, temperature, sample count and acceptance. Record force, displacement and electrical behavior on a common sample identity where useful. Inspect hidden interfaces through the approved method after conditioning.

Do not use a single room-temperature pull test to claim dynamic-flex life. The dynamic-flex harness RFQ guide covers motion profile, service loops and cycle evidence. Likewise, do not treat a pull pass as an IP rating; the sealed-connector leak-test guide explains boundary and correlation requirements.

Compare three fictional supplier packages

The following packages are invented for procurement training. They are not customer cases or supplier performance claims.

Supplier A returns a color photograph and says “100% waterproof overmold.” The cable jacket material is unknown, no section is shown, and the report contains only continuity. This package is insufficient for adhesion, cure, void, sealing and strain-relief approval. Request the controlled BOM, boundary, process and evidence.

Supplier B identifies the compound and cable, provides a TDS and records dispense weight. Its samples pass an axial pull at room temperature. However, no mix-ratio record, hidden-void method, failure mode or post-conditioning result is supplied. The package is partly reviewable. Keep the missing functions open rather than rejecting unrelated evidence.

Supplier C freezes every substrate and material, supplies section drawings and process windows, links each shot to metering and cure records, and provides initial plus conditioned samples with void images, force/displacement traces and post-test electrical/leak checks. This is the strongest package because the buyer can compare it with the project. It is not automatically acceptable: its sample matrix, conditioning, detection capability and limits still require technical approval.

The comparison shows why a large headline pull force can be less useful than a complete configuration and failure-mode record.

RFQ evidence-request matrix

On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.

Build a representative sample matrix

Select samples around actual risk. Consider:

• minimum, nominal and maximum cable jacket diameter;
• each approved jacket polymer, color and compound/source variant;
• each connector housing or backshell resin and mold cavity where relevant;
• thinnest and thickest overmold/potting section;
• most congested wire arrangement and smallest vent path;
• straight, angled and branched cable exits;
• highest and lowest approved material/component temperatures;
• shortest and longest allowed wait after surface treatment;
• beginning and end of material pot life or residence window;
• each tool cavity, nozzle, mixer or dispense head;
• start-up, restart after interruption and steady production;
• initial and environmentally conditioned assemblies; and
• manual versus automated process variants, if both are allowed.

Use released production tooling and operators. A laboratory casting in an open cup cannot bound an injected overmold around a tightly packed connector. Record deliberate worst cases and the reason they represent the production range.

If a destructive section is used, define planes by datums and distribute sections across critical zones. One attractive cross-section cannot establish complete three-dimensional fill. If CT is used, correlate image indications with sections or known references until the method’s detection boundary is understood.

Put measurable production controls around the process

Incoming material control

Verify supplier, product, lot, packaging, expiration, storage and transport condition. Record thaw/conditioning time where required. Quarantine damaged, separated, crystallized, contaminated or out-of-window material under the manufacturer’s rules. Do not extend shelf life without approved evidence.

Setup and first-off control

Verify tool and cavity, fixture, nozzle/mixer, purge, temperatures, pressure/dispense settings and cable/insert position. Section or otherwise evaluate first-off pieces at the defined frequency. Keep start-up scrap separate from approved product.

In-process control

Capture ratio or pump performance, material temperature, shot/dispense amount, cycle/timers, vacuum/vent status and alarms. Use process data to prevent escape; do not present it as proof of hidden quality unless correlation has been established.

Cure and release control

Link assemblies to mix/shot and cure time/temperature. Prevent packing, electrical test, pulling or bending before the approved release state. Record rework or interrupted cure and its disposition.

Final and periodic control

Check dimensions, flash, short shot, cable position, surface defects, markings and mating surfaces. Run the defined periodic section/CT, adhesion, pull/torsion/flex, leak and electrical checks. A failed periodic check needs a time/lot boundary, containment, investigation and retest plan.

NASA-STD-8739.1B’s mix and material traceability rules illustrate the value of connecting a material batch to the hardware where it was used. Apply the project’s own contract requirements and records; do not label a commercial process “NASA compliant” without formal applicability and evidence.

Control rework explicitly

Overmold and potting rework can be more risky than initial production because cured material may be difficult to remove without nicking insulation, changing a seal, bending a terminal or contaminating a new bond surface.

The RFQ should state:

• whether rework is prohibited, restricted or allowed;
• which defects can be reworked;
• approved removal tools and damage limits;
• whether terminals, housings, wires or seals must be replaced;
• preparation of the old cured interface;
• compatible repair material and cure;
• required dimensional, electrical, leak and mechanical repeat checks;
• serialized rework history; and
• maximum permitted rework cycles, if any.

Do not hide a void with a cosmetic surface patch unless the approved engineering disposition shows that the repair reaches the critical zone and restores every required function. Do not ship a destructively sectioned or pull-tested specimen.

Follow a seven-step procurement decision

1. Freeze the application and failure consequence

Define motion, handling, water/fluid, temperature, voltage, contamination, UV and service exposure. Identify whether the polymer is required for strain relief, sealing, insulation, thermal management or several functions.

2. Freeze the complete interface BOM

Approve connector/housing, terminals, wire/cable, jacket compound, seals, pre-mold, polymer system, primer and cleaner. Record variants and tolerances.

3. Approve the section and load-path drawing

Show material boundaries, mechanical interlocks, fill/vent route, cable exit, critical zones and installation support. Resolve where axial, torsional and bending loads travel.

4. Approve the material and process window

Review storage, preparation, ratio, mix, pot life, temperatures, pressure, fill, vacuum/vent, fixture and cure. Link each parameter to a record and reaction rule.

5. Approve the qualification and inspection plan

Select samples, conditioning, adhesion/retention methods, void detection, sealing/electrical checks, acceptance criteria and failure analysis. Keep each function separate.

6. Validate production and first articles

Build with released production equipment and traceability. Review sections/images, process data, mechanical traces and post-test results against the exact offered assembly.

7. Release routine control and change rules

Set first-off, inline, final and periodic checks, failed-check containment, record retention and requalification triggers. Confirm packaging and installation keep loads within the approved boundary.

Changes that should trigger review

Require advance notification for changes to:

• connector housing, backshell, seal, terminal, splice or pre-mold;
• wire/cable supplier, jacket or insulation compound, color, OD, conductor or lubricant;
• overmold/potting manufacturer, product, hardener, primer, cleaner, formulation or package size;
• material lot-control, shelf-life extension, storage, thaw or conditioning practice;
• surface preparation, plasma/corona/primer treatment or treatment-to-use time;
• meter/mix equipment, pump ratio, mixer, nozzle, purge or batch size;
• tool steel, mold cavity, gate, vent, insert fixture or cooling circuit;
• injection/dispense temperature, pressure, speed, shot amount, vacuum or timing;
• cable position, branch angle, free length, support or strain-relief geometry;
• cure schedule, oven, fixture-release rule or witness-sample method;
• inspection method, CT/X-ray settings, section plane, algorithm or reviewer criteria;
• mechanical/leak/electrical test fixture, method, limit or sample frequency;
• factory, line, automation, operator qualification or subcontractor; and
• drawing, customer standard, IPC/WHMA-A-620 or other invoked requirement revision.

Classify each change by affected interface and function. A cable color change may also change the jacket formulation. A nozzle change may alter mixing, air inclusion and fill pattern. A mold repair can alter venting. “Same nominal material” is not an equivalence decision.

Send a complete overmold or potting RFQ

Provide the application environment, connector and harness BOM, drawings, cable route/support, declared polymer functions, critical-zone map, process/evidence requirements, sample quantities, qualification sequence, production records and change-control expectations. Ask bidders to return exact materials, current manufacturer documents, section/load-path drawings, validated process windows, inspection capability, applicable reports, deviations and separate commercial terms.

Request price, tooling, sample cost, MOQ and lead time for the defined configuration in writing. No commercial term follows from a compound family, process name or this guide.

Send a cable-harness overmold and potting RFQ with the interface BOM and section drawing. Ask for evidence that separates adhesion, cure, hidden fill, sealing and strain relief, plus a clear list of conditions that remain unverified.

Buyer FAQ

Is overmolding the same as potting?

Not necessarily. Overmolding commonly forms material around an insert in a mold, while potting commonly dispenses a flowable compound into a cavity to cure. Suppliers may use terms differently, so define the process, materials, geometry and evidence on the drawing.

Does a full-looking mold prove adhesion?

No. The shape can be retained by mechanical interlock while an interface is unbonded. Define which interface must adhere, the preparation, test method, conditioning, acceptance and allowed failure mode.

Does a strong axial pull prove good strain relief?

Only for the tested direction, fixture, cable, cure state and acceptance boundary. Confirm that the jacket-to-body load path engages before terminal retention, and separately evaluate torsion, lateral bending or repeated flex where those occur.

Can continuity prove that potting is acceptable?

No. Continuity can remain normal despite hidden voids, weak adhesion, off-ratio material or a termination carrying unintended mechanical load. Use function-specific inspection and tests.

What is the right void percentage?

There is no universal percentage for every overmold. Define critical zones and failure consequences, then derive limits from the governing design, standard and validated process. A connected void path can matter more than the same total isolated volume elsewhere.

Should every compound be vacuum degassed?

Follow the exact material instructions and validated process. NASA-STD-8739.1B requires degassing within its encapsulation scope. Another chemistry or dispensing system may require a different method, and the current manufacturer instructions must govern the selected compound.

Can shot weight replace X-ray or sectioning?

No. Weight can show that the total amount is plausible but cannot locate a hidden void, misplaced wire or unbonded interface. It can be one process control when correlated with a more direct inspection.

How should cure be verified?

Use the material manufacturer’s instructions and a validated release rule linked to the hidden critical region. Time/temperature history, a same-batch witness sample and a correlated property check may be combined. Surface feel alone is generally insufficient evidence.

Does a TDS value become the drawing acceptance limit?

Not automatically. Manufacturers often label data as typical or representative. Identify which value is a specification, which is process guidance and which project acceptance requirement has been validated on the complete assembly.

Can the supplier change to another black TPE or epoxy?

Only after documented review. Chemistry, additives, viscosity, shrinkage, hardness, adhesion, cure and environmental behavior may differ. Repeat the justified compatibility, process and qualification work before release.

What should happen after a failed periodic pull or void check?

Stop or contain production under the approved reaction plan, identify product since the last known-good check, investigate material/process/test causes, disposition affected units and demonstrate restored control before release.

Can a potted connector still leak?

Yes. A continuous path can exist along a housing interface, wire jacket, strands or incompletely filled region. Define the boundary and use a correlated leak or ingress method; a visible fill is not a leak result.

Should samples be tested after environmental conditioning?

When the application risk requires it, yes. Use the same serialized samples and a written sequence where feasible. Heat, moisture, fluid, vibration or thermal cycling can change adhesion, material properties and residual stress.

Does this guide confirm that SINAWATTS can overmold or pot a harness?

No. It is a procurement framework. Request current written capability, exact materials/equipment, process scope, samples, evidence and commercial terms for the proposed assembly.

Official sources checked on 2026-10-04

• NASA-STD-8739.1B with Change 2 official record and official PDF — active NASA polymeric-application scope, mix traceability and encapsulation workmanship requirements.
• IPC/WHMA-A-620F official table of contents and IPC A-620 public scope — current cable/harness acceptance scope, potting/over-molding sections and the limits of generic inspection-frequency and hidden-feature criteria.
• Molex custom cable assembly solutions — manufacturer description of overmolded strain relief at the connector/cable interface.
• 3M Scotch-Weld DP270 Clear May 2026 TDS — exact-product potting/encapsulation, surface preparation and typical-data boundaries.
• DOWSIL TC-6020 official product page — exact-product identity, mixing ratio, cure conditions and typical-property boundaries.
• IEC 60512-17-3:2010 official record and IEC 60512-17-4:2010 official record — cable-clamp tensile and torsion test-method scopes when invoked by a detail specification.

These sources were used only within their published scopes. None proves suitability, certification or production capability for an unnamed cable harness or supplier.