When checking performance in prototyping, if assembly is done in a way that differs from actual volume production conditions, the evaluation results may not carry over to volume production.

For example, conditions such as

・Applying slowly by hand

・Compression bonding more strongly than usual

・Excluding prototypes with air bubbles

・Measuring after allowing a long time to settle

・Using only materials in good condition

・Fixing ideally with jigs

Temperature results obtained this way are valid for checking the potential of the material.

However, they do not necessarily show the real capability of volume production equipment.

At the stage of judging whether to move to volume production, it is necessary to reproduce as closely as possible

・The actual processing speed

・The actual tension

・The actual compression bonding time

・The actual application equipment

・The actual storage time

・The actual peeling conditions

The purpose of prototype evaluation is not to prove maximum performance.

It is to confirm that the required performance is achieved even under volume production conditions.

When deciding volume production conditions, evaluating only the best conditions is not enough.

In actual volume production, conditions fluctuate little by little.

That is why it becomes important to deliberately combine and evaluate

・Upper and lower limits of material thickness

・Upper and lower limits of compression bonding pressure

・Upper and lower limits of lamination speed

・Changes in temperature and humidity

・Die wear states

・Conditions with high and low peeling force

By varying conditions in this way, it is possible to check

・Which factors strongly affect thermal performance

・How far conditions can change while maintaining performance

・Which combinations cause defects

What is needed in volume production is not an optimal condition at a single point.

It is a processing window in which performance does not collapse even if conditions shift somewhat.

Volume production conditions have allowable upper and lower limits.

For example, with compression bonding pressure,

if too low, air bubbles remain.

If too high, the material is crushed.

In between lies a range in which processing can be done stably.

In volume production, conditions close to the limits should not be made the standard.

For example,

just barely the minimum required pressure.

Just barely the maximum pressure at which the material does not break.

Just barely the equipment’s maximum speed.

Under such conditions, small fluctuations cause defects.

It is better to place standard conditions

near the center of the allowable range.

This is because even if process conditions fluctuate somewhat, they are less likely to exceed the upper or lower limits.

The same applies to thermal management components.

Rather than the single point that gives the best temperature performance, it is important to choose a range that is stable in volume production.

Just because a product’s dimensions and quality items once fell within specification does not mean the process is stable.

In volume production, they must continuously fall within specification.

When looking at process capability, check

・Whether the average value is near the center of the specification

・Whether variation is sufficiently small relative to the specification width

・Whether the average value is drifting over time

・Whether there are changes in specific lots or time periods

For example, even if all outer dimensions are within specification, if the values are concentrated near the upper limit, a small change may push them out of specification.

Even if burr height is currently within specification, if it is rising with the number of processed pieces, it may soon turn into defects.

The purpose of looking at process capability is not to confirm that things passed in the past.

It is to judge whether things can continue to pass stably in the future.

Dimensions and thickness are relatively easy to measure.

On the other hand, measuring thermal resistance or temperature on every piece may be difficult in terms of time and equipment.

So in volume production, substitute characteristics related to thermal performance are sometimes controlled.

For example,

・Thickness

・Contact area

・Air bubble area

・Lamination position

・Burr height

・Compression ratio

・Peeling force

・Flatness

・Adhesive layer thickness

However, controlling these does not necessarily guarantee thermal performance.

What matters is checking at the development stage

which processing characteristics are strongly related to thermal performance.

For example,

as air bubble area increases, temperature rises.

When thickness exceeds a certain range, thermal resistance increases.

When the compression ratio is low, contact deteriorates.

If these relationships can be confirmed, then in volume production, abnormalities become easier to detect by controlling the important processing characteristics, even without measuring thermal performance directly.

When image inspection equipment is introduced, many items can be measured.

Outer dimensions.

Hole positions.

Chipping.

Foreign matter.

Air bubbles.

Misalignment.

However, it is not the case that every measurable item should be strictly controlled.

If inspection items increase too much,

・False detections increase

・Equipment stoppages increase

・Judgment becomes complicated

・Truly important abnormalities get buried

・Costs rise

may result.

What matters in inspection design is

prioritizing items that affect performance and safety.

For example, a minute lift directly above the heat source may be more important than a small stain that is visually conspicuous.

Exposure of conductive material is more important than slight discoloration at an edge.

Inspection is not done only to select clean-looking components.

It is done to find abnormalities that break function.

Inspecting every item 100% may not be realistic.

On the other hand, it is also difficult to leave serious defects to sampling inspection alone.

That is why inspection methods must be divided by item.

For example,

Items suited to 100% inspection

・Presence or absence of the product

・Large misalignment

・Missing parts

・Clear outer shape defects

・Large air bubbles

・Exposure of conductive material

・Component pitch

・Front/back and orientation

Items suited to sampling inspection

・Thickness

・Peeling force

・Burr height

・Adhesive strength

・Fine dimensions

・Thermal performance

・Material properties

However, which items to inspect 100% must be decided by considering

・Defect occurrence frequency

・Impact on customers

・Detectability

・Process stability

・Inspection cost

Air bubbles, wrinkles, misalignment and the like can sometimes be checked by images.

On the other hand,

・Insufficient contact inside the adhesive layer

・Fine interlayer delamination

・Cracks inside graphite

・Thickness changes after compression

・Local insufficient contact pressure

are difficult to check by appearance alone.

Good appearance does not necessarily mean good thermal performance.

That is why, at the development stage, it is necessary to combine

・Cross-section observation

・Thickness measurement

・Peel testing

・Compression testing

・Thermal resistance measurement

・Temperature distribution measurement

・Temperature cycle testing

to confirm the relationship between appearance and function.

Once the relationship is confirmed, appearance and dimensions can sometimes be used as substitute indicators in volume production.

However, guaranteeing thermal performance by appearance alone, without confirming that relationship, is difficult.

To check product variation, measurement must be done correctly.

However, measured values also differ depending on

・The person measuring

・The measuring equipment

・The measurement position

・The measurement load

・The measurement time

・The ambient temperature

・How the item is placed on the jig

For example, when measuring the thickness of soft TIM, results change if the probe load is different.

In thermal evaluation, temperature results change if

・The output of the heat-generating body

・Sensor position

・Fastening force

・Ambient temperature

・Time until the start of measurement

differ.

A difference that looked like product variation may actually be measurement variation.

Therefore, in volume production control, before aligning the products, align the measurement methods.

Even with the same material name and the same product number,

・Thickness

・Hardness

・Surface condition

・Adhesive strength

・Peeling force

・Friction

・Moisture content

may change slightly depending on the material lot.

Even if these changes are within the material specification, their effect on processing may differ.

For example,

with a lot on the high adhesive strength side, peeling is difficult.

With a lot on the hard side, chipping occurs easily during die-cutting.

With a lot on the thick side, the amount of compression increases.

With a lot on the rough surface side, air bubbles remain easily.

That is why, before moving to volume production, it is better to check processability with multiple lots if possible.

If conditions are decided with only one material lot, they may become narrow processing conditions that suit only that lot.

Materials are sometimes changed due to supply problems, cost, discontinuation and so on.

Even if the material manufacturer explains it as an “equivalent product,” the processing conditions are not necessarily the same.

Even if thermal conductivity and thickness are equivalent,

・Adhesive strength

・Peeling force

・Hardness

・Surface friction

・Elongation

・Die-cuttability

・Powder shedding

・Warping

may differ.

In volume production processing, small differences in properties affect

・Die conditions

・Lamination pressure

・Tension

・Peeling speed

・Ease of scrap removal

Therefore, when changing materials, do not judge by the material data sheet alone;

it is important to check with the actual laminated structure and volume production conditions.

Even if equipment of the same model is used, processing results may change if

・Roller condition

・Installation accuracy

・Sensor calibration

・Die mounting state

・Transport path

・Temperature distribution

・Control response

differ.

Conditions that were stable at a factory in Japan will not necessarily give the same quality if transferred as-is to an overseas factory.

Rather than transferring only equipment names and set values, it is necessary to check

・Actual pressure

・Actual tension

・Actual speed

・Material condition

・Characteristics of the finished product

In process transfer, the purpose is not copying the condition sheet,

but reproducing the same product state.

In processes that are not automated, differences between workers can affect quality.

For example, in hand application,

・The position where application starts

・The application angle

・Compression bonding force

・Application speed

・Number of reapplications

change depending on the worker.

Skilled workers unconsciously let air bubbles escape and correct component warping.

However, if those motions are not standardized, other workers cannot reproduce the same results.

In moving to volume production, it is necessary to clarify

・Work procedures

・Jigs

・Compression bonding conditions

・Judgment criteria

・Training methods

A state in which “it cools if a skilled person makes it” cannot be called stable as a volume production process.

In prototyping and manual work, a misaligned component is sometimes peeled off and reapplied.

Even if it looks normal again,

・The adhesive may stretch

・Adhesive strength may change

・Foreign matter may get in

・Unevenness may remain on the adhesive surface

・The component may deform

Reapplied products and first-application products do not necessarily have the same contact state or thermal performance.

Therefore, it is necessary to decide

・Whether to allow reapplication

・How many times to allow it

・What the conditions for reuse are

・Whether re-evaluation is needed

Components right after lamination and components stored for several days may be in different states.

Adhesive may conform to the mating surface over time after bonding.

On the other hand, peeling force may change as winding tightens during reel storage.

Materials may absorb moisture, changing dimensions and softness.

That is why evaluation conditions change depending on whether

・It is used immediately after processing

・It is stored for a week

・It is stored for several months

Volume production specifications may need to include

・Storage temperature

・Storage humidity

・Expiration date

・Usable time after opening

・Material settling time

Products are not necessarily used only inside the factory.

After completion, they reach customers via

・Trucks

・Ships

・Air freight

・Warehouses

During transport, they may be subjected to

・High temperatures

・Low temperatures

・High humidity

・Vibration

・Shock

・Long-term stacking loads

Reels shift sideways.

Components move on the carrier.

Adhesive flows.

Winding curl becomes stronger.

Foreign matter enters from the outer packaging.

Such problems may occur.

Therefore, the quality of volume production components must be checked not right after manufacturing,

but at the point of arrival at the customer.

Packaging and transport conditions are also part of process design for maintaining processing quality.

In volume production, it is important not only to bring defects completely to zero, but also to be able to identify the scope when an abnormality occurs.

For that, make it possible to trace

・Raw material lot

・Processing date and time

・Equipment used

・Die used

・Processing conditions

・Inspection results

・Worker

・Reel number

・Shipping destination

For example, if a temperature abnormality is found at the customer, being able to check

whether it is the same material lot,

whether it is the same period of die use,

whether it is the same position on the reel,

whether it is the same processing time period,

makes it easier to advance the cause investigation.

If it cannot be traced, it becomes necessary to stop or recall a wide range to be safe.

Traceability is not only for after a quality problem occurs.

It is also needed

to learn the causes of volume production variation and improve the process.

Investigating only the products that had problems sometimes does not reveal the cause.

What matters is comparing normal products and abnormal products.

For example,

abnormal products had a high number of die uses.

Abnormal products were concentrated at the start of reel winding.

Abnormal products increased with a specific material lot.

Abnormal products were processed on high-humidity days.

This comparison also requires records on the normal product side.

Therefore, it is important not to collect process data after problems occur,

but to keep it from normal times.

In volume production control, finding out-of-specification products is important.

However, better control is finding changes before things go out of specification.

For example, changes such as

・Burr height gradually increasing

・Peeling force trending upward

・Air bubble area increasing little by little

・Component pitch shifting in one direction

・Lamination position approaching the upper limit

Even if things are currently within specification, looking at trends may make it possible to predict that defects will occur in the near future.

If at this stage it is possible to

・Clean the die

・Replace blades

・Adjust tension

・Inspect rollers

・Check material condition

it becomes easier to prevent defects from occurring.

In volume production quality, what matters more than sorting out defective products

is stopping changes in the direction that produces defects.

Automated inspection using image inspection and sensors is sometimes described as a means of labor saving.

In thermal management components, however, that is not the only purpose.

With the human eye,

・Judgment criteria change

・Oversights increase due to fatigue

・Minute positional differences are hard to quantify

・Long-term trends are hard to grasp

With automated inspection, it is possible to

・Judge by the same criteria

・Record as numerical values

・Check changes over time

・Find abnormal trends early

・Link to product history

In other words, the essence of automated inspection lies not in reducing the number of inspectors,

but in continuing to record processing states related to thermal performance by the same criteria.

Image inspection is effective, but it is not all-powerful.

What can be seen in images is mainly surfaces and outlines.

Internal contact states.

Air bubbles inside the adhesive layer.

Fine interlayer delamination.

Cracks inside materials.

Actual thermal resistance.

These may not be directly checkable.

That is why it is necessary to combine

・Image inspection

・Dimensional measurement

・Peel testing

・Cross-section checks

・Thermal evaluation

・Environmental testing

Automate the items that can be automated.

However, keep periodic evaluation for items that require functional checks.

What matters is not increasing inspection methods.

It is clarifying which method detects which abnormality.

When prototyping succeeds, one wants to increase volume production quantities quickly.

In initial volume production, however, problems that were not visible in prototyping appear, such as

・Die wear

・Rising equipment temperature from long operation

・Tension changes in the second half of the reel

・Material lot differences

・Worker changes

・Effects of storage time

・Accumulation of scrap and adhesive

That is why volume production start-up requires stages such as

1.Check process conditions in small quantities

2.Observe changes in continuous processing

3.Check with different material lots

4.Evaluate after storage and transport

5.Raise production speed

6.Standardize the stable range

The purpose of volume production start-up is not to produce quantities as fast as possible.

It is to find conditions under which quality does not collapse even with long continuous production.

If the defect rate is low right after the start of volume production, the process looks good.

In reality, however, it may be that

・Defects are not being found by inspection

・Specifications are not linked to function

・Only good material lots are being used

・The die is new

・Continuous operation time is still short

Conversely, even if there are many defects in early volume production, if the causes are identified and process conditions improved, things may stabilize afterwards.

Therefore, what should be looked at is not only the defect rate.

What matters is

・What the cause was

・Whether the cause could be reproduced

・Whether recurrence prevention was built into the process

・Whether it can withstand changes in conditions

・Whether trends can be monitored

In volume production processes, various changes occur.

・Change of material manufacturer

・Change of material manufacturing site

・Change of material thickness

・Change of release film

・Die renewal

・Equipment change

・Change of processing speed

・Change of packaging method

・Change of production factory

Even if the change looks small, it may affect thermal performance and automated supply.

For example, merely changing the release film can change peeling force and increase component deformation.

Merely raising processing speed can change how air bubbles escape.

That is why, when changes are made, it is necessary to check not only what changes,

but also which functions the change may affect.

Evaluation that reproduces volume production conditions.

Checking weak points by varying conditions.

Understanding process capability.

Using inspection items appropriately.

Standardizing measurement methods.

Checks when material lots change and when equipment is transferred.

Dealing with variation in work done by people.

The effects of storage and transport environments.

Securing traceability.

A staged approach to volume production start-up.

And thorough change management.

All of these are efforts to turn “it cooled once” into “it keeps cooling.”

Each may look like a modest control item.

However, without this accumulation, prototype results cannot be carried over to volume production.

In the final part, we organize what has been covered so far into eight steps.

Then, after checking the perspectives of researchers, designers, the manufacturing floor, quality departments, and sales and development, we finally describe the role OTIS plays.

 

Column Supervisor: Koji Kakumoto (President, OTIS Co., Ltd.)

After studying abroad and working in planning and development at a trading company, he joined OTIS Co., Ltd. in 2011. While primarily working in the Corporate Planning Department, he has also served concurrently in manufacturing and technical divisions, and since 2018 has served as President, working to drive business growth and strengthen the organization.

 

This article is a general technical explanation intended for educational purposes and does not refer to any specific company, product, or technology.