How Should Buyers Prove Dispensing Process Equivalence at a Replacement Supplier?

Home / Blog / Replacement supplier process equivalence

How Should Buyers Prove Dispensing Process Equivalence at a Replacement Supplier?

Process equivalence is often reduced to a screenshot comparison: same ratio, same pressure, same path speed. It feels precise. It can also be wrong.

A new supplier may use a different pump, valve, hose length, mixer, fixture or controller. The same displayed setting can create a different deposit. Conversely, different settings can produce the same conforming bead and cured function. Buyers should compare the physics and the product result, not the appearance of the recipe screen.

The six-layer equivalence model

Layer	Question	Typical evidence
Material	Is the chemistry and condition approved?	Product code, batch, ratio, age, storage, temperature
Equipment	Can the architecture control the material?	Pump/valve/mixer/fixture review and alarms
Deposit	Is the uncured output equivalent?	Mass, flow, bead width/height, position, fill level
Function	Does the cured part perform?	Void, hardness, adhesion, leak, dielectric or thermal tests
Capability	Can the process repeatedly meet limits?	Distribution, stability and suitable capability measures
Production	Does it hold across real variation?	Shifts, lots, startups, stops, refill and rate

ISO 22514 describes capability and performance studies used to judge whether manufacturing output or equipment is acceptable against suitable criteria. Acceptance sampling standards can support lot decisions, but neither statistics nor sampling rescues a weak test method. Start with meaningful requirements.

Layer 1: lock the material definition

Confirm manufacturer, product code, component ratio, approved alternatives, filler content, shelf life, storage and conditioning. If the material itself changes, the project is no longer a pure process transfer. Treat material equivalence and process equivalence as separate approvals.

Record actual production temperature and time out of storage. A highly filled thermal compound at 20°C can load a pump differently than at 28°C. “Same viscosity” from a data-sheet headline may not describe shear behavior or lot variation well enough.

Layer 2: compare equipment by function

Element	Functional comparison	Transfer risk
Metering pump	Flow range, ratio control, filler tolerance	Pulsation, wear or component imbalance
Valve/nozzle	Response, shutoff, pressure and geometry	Drip, stringing or bead change
Static/dynamic mixer	Mixing energy, volume and residence time	Streaking or cure variation
Motion/fixture	Path accuracy, datum and repeatability	Misplaced bead or overflow
Material supply	Heating, agitation, degassing and refill	Air, settlement or viscosity drift
Control system	Recipe security, alarms and traceability	Undetected or unauthorized change

The comparison should explain how each system controls the risk. It does not need matching brand names.

Layer 3: compare the deposit before cure

Measure what the machine actually delivers. For dots and shots, deposited mass is often more useful than programmed time. For beads, measure width, height, continuity, position and start/stop behavior. For potting, include total fill mass, fill height, overflow and visible or internal air.

Use the same calibrated method at both sources. If two laboratories use different scales, image thresholds or section locations, apparent differences may come from measurement rather than process.

Layer 4: prove cured product function

Application	Product-level evidence	Hidden risk
EV battery potting	Void, dielectric, thermal path and cure	Internal air or poor interface contact
PCB adhesive	Coverage, adhesion, contamination and cure	Keep-out violation or weak bond
LED driver encapsulation	Fill, hardness, electrical and aging	Incomplete cure below surface
Sensor gasketing	Bead continuity, compression and leak	Seal discontinuity after assembly
Industrial bonding	Lap shear/peel or agreed functional test	Surface-dependent adhesion loss

Choose tests from the product function. A beautiful bead is not equivalent if the assembly leaks. A matching hardness is not enough if thermal resistance changes.

Layer 5: compare capability only after checking stability

Capability indices are tempting because they compress data into one number. Use them carefully. The process should be stable, the distribution and subgroup logic should be understood, and specification limits should represent product needs. Report the raw trend and sample context alongside any index.

Do not declare equivalence because both processes exceed a chosen Cpk target on one short run. Include centering, spread, outliers, start-up behavior and measurement uncertainty. For destructive tests or expensive assemblies, combine a justified sample plan with prevention controls and process monitoring.

Layer 6: expose normal production variation

Run normal operators, intended takt time and realistic duration. Include cold start, restart after a pause, material refill, mixer change, shift handover and at least one material-lot transition where feasible. These moments often reveal differences hidden by a continuous demonstration run.

Variation source	Comparison question	Failure signal
Material lot	Does output stay inside the window?	Mass, ratio or cure shift
Operator/shift	Is setup independent of one expert?	Shift-linked defects
Restart/refill	Are purge and first-piece rules effective?	Air, short shot or ratio transient
Maintenance	Are wear limits and replacements defined?	Gradual bead or pressure drift
Production rate	Can the system sustain takt?	Heat, residence-time or capacity issue

Use a pre-agreed equivalence decision matrix

Result	Decision	Action
All critical requirements met; stable evidence	Equivalent	Proceed to phased launch
Product meets requirements; long-term data limited	Conditionally equivalent	Limit volume and retain launch controls
Difference understood and beneficial	Approve documented deviation/change	Update specifications and validation
Critical function or control fails	Not equivalent	Redesign, retest or reject source
Measurement cannot distinguish processes	Inconclusive	Improve test method before deciding

Make sure the measurement systems are comparable

Two suppliers can produce identical parts and still report different numbers when their measurement systems disagree. Before interpreting a process gap, align the scale resolution, image-analysis threshold, cross-section location, conditioning time, test speed and fixture method. Where practical, send a common set of reference parts to both sites and compare results without revealing the expected answer.

For deposit mass, verify balance resolution, tare practice, evaporation or cure delay and the number of deposits combined in one reading. For bead geometry, define where width and height are measured and how start/stop regions are treated. For void analysis, agree whether the result is based on area, volume, maximum single void or total void fraction. Functional tests need the same specimen preparation and environmental conditioning.

If measurement variation is large relative to the product tolerance, the study cannot confidently distinguish the processes. Improve the method before expanding sample size. More data from a weak gauge only makes the uncertainty look more formal.

Separate statistical difference from engineering importance

A large sample can identify a small numerical difference that has no practical effect. A small sample can miss a difference that matters greatly near a functional limit. Review both statistical evidence and engineering margin.

For example, a replacement process may center deposit mass 1% higher than the old source while remaining comfortably inside the qualified fill and stress window. That may be acceptable. A similar 1% shift in mix ratio could be unacceptable for a chemistry with a narrow cure window. The decision should consider specification margin, failure mechanism, detectability and downstream function.

Document known differences rather than forcing a false claim that the processes are identical. Equivalence means the replacement process satisfies the approved requirements with comparable or controlled risk. It does not mean every distribution, machine component or operating value must be indistinguishable.

Equivalence approval checklist

• Same approved drawing, material and functional requirements used
• Equipment differences mapped to process risks
• Measurement systems and definitions aligned
• Paired or matched samples used where practical
• Uncured deposit and cured function both evaluated
• Critical defects have explicit accept/reject rules
• Stability reviewed before capability claims
• Multiple runs, shifts, lots and transitions represented
• Residual differences documented and approved
• Launch controls and rollback conditions defined

Frequently asked questions

Do the old and new dispensing machines need identical settings?

No. Settings are equipment-specific. Equivalence should be based on approved material, deposit characteristics, cured performance, process capability and stability under representative production conditions.

What is the best first comparison between suppliers?

Use paired parts or matched lots where practical. Apply the same drawing, material specification and inspection method, then compare deposited mass, bead geometry, voids, cure and functional performance.

Is a visual comparison enough?

Rarely. Visual inspection can support bead and surface checks, but hidden voids, ratio drift, incomplete cure, weak adhesion, leakage or thermal performance require suitable measurements or functional tests.

Must the replacement process outperform the old process?

Not necessarily. It must meet approved requirements with acceptable risk and capability. A buyer may set tighter targets when the transfer is intended to correct a known weakness.

How many samples prove equivalence?

There is no universal number. Sample size should reflect variation, destructive-test cost, risk and confidence required. Cover multiple runs, operators, material lots and process transitions rather than one convenient batch.

When can equivalence be approved conditionally?

Conditional approval is reasonable when product requirements are met but long-term production evidence is still developing. Limit volume, apply launch controls and define the evidence needed for normal release.

Approve equivalent results, not matching screen values

A sound equivalence study explains why the replacement system can repeatedly create the required product. If your team needs sample testing or a replacement dispensing configuration, send OBO Precision the material data sheet, part drawing, old process information, output target and acceptance criteria.

Request an Engineering Review

Related guides: production transfer after supplier removal, production validation run, and launch control plan.

Related inventory guide: see how to plan last-time-buy inventory during a dispensing supplier transfer.

Related customer approval guide: prepare a customer approval package for a dispensing supplier transfer before implementing or shipping changed production.

Related cutover guide: use this first-shipment cutover readiness checklist for a replacement dispensing supplier.

Next control step: After the replacement source is approved, use a documented receiving inspection plan for its first commercial shipments before unrestricted production release.

Disposition follow-up: If first-shipment inspection fails, use a controlled conditional-release decision for the replacement dispensing supplier rather than an informal use-as-is approval.

Continuity follow-up: When a failed first shipment must be rejected, use a documented production recovery plan for the replacement dispensing supplier instead of weakening the release criteria.

Recovery-lot validation: Before resuming supply, require focused validation of the expedited replacement lot, including change review, targeted testing and controlled line release.

Finished-goods follow-up: After a recovery lot enters production, use a controlled finished-goods release plan before authorizing customer shipment.

Post-shipment follow-up: After releasing recovery-lot finished goods, use a defined post-shipment monitoring and customer containment plan to detect recurrence early.

Complaint-response follow-up: If the first customer issue appears, use a structured recovery-lot complaint and containment process before calling it isolated.

Customer containment: When affected stock remains at the customer, use a validated customer stock screening plan rather than relying on an unproven visual sort.