Choose the mold that controls the part

Generated from content/lms/fabrication-and-composites/03-make-tools-that-control-shape/02-choose-open-mold-or-matched-mold.md; edit the source file, not this page.

Source path: content/lms/fabrication-and-composites/03-make-tools-that-control-shape/02-choose-open-mold-or-matched-mold.md

Course: Fabricate composite race-car parts with workshop discipline

Module: Make tooling that controls the finished part

Estimated duration: 55 minutes

Skill aim

You are choosing the tool before you choose the layup day. The skill is not to admire matched molds or to default to the fastest open mold. The skill is to decide which surfaces, dimensions, fibers, cure conditions, and production count actually need control, then choose the least complicated tool that gives that control without creating a new failure mode.

In this lesson, open mold means a single working mold surface, usually a female mold for bodywork-style parts, where the visible face of the part is controlled by the mold and the back face is controlled by your layup workmanship, peel ply or release material, trimming, and finishing. Matched mold means two working mold surfaces, male and female, used together so the laminate is pressed between them. In between those two poles are useful pressure-assisted methods: a flat board over polythene sheet, a sandbag on a simple curved honeycomb panel, or vacuum consolidation over an open mold. Those halfway methods matter because many race-car parts do not justify a true matched tool, but they still need more control than hand contact alone.

The governing rule

Choose open mold when one face defines the part, the back face can be rougher or trimmed, hand access is good, the shape can be consolidated by contact pressure, and the number of parts does not justify a second tool face. Choose matched mold when the back face also matters, laminate thickness must be repeatable, the laminate needs stronger consolidation than hand pressure, carbon fabric is reluctant to sit into tight curvature, or you will make enough parts that extra tooling time pays back in fewer defective parts.

The mechanism is simple. A composite part is a matrix and reinforcement working together. The mold decides where the laminate is allowed to cure. The pressure method decides how well the reinforcement, resin, and any core stay in contact with that shape. If you only control one side, the other side will reflect the release film, hand compaction, fabric overlaps, resin richness, and trimming decisions. If you control both sides, you can make thickness and the inner surface much more consistent, but only if the two halves are designed so they close cleanly, separate cleanly, and apply pressure where the laminate actually needs it.

This is why the mold decision is a control decision, not a prestige decision. A home workshop can make useful GFRP parts, carbon parts, aramid parts, sandwich panels, body panels, spoilers, aerofoils, ducting, dashboards, and other competition-car components. But every extra control layer costs time, materials, cure management, release work, and demolding risk. The correct question is not whether matched molding is better. The correct question is what the part cannot be allowed to vary.

Start from the part function

Before you sketch the tool split, write the part function in plain language. A nosecone or duct skin needs an exterior shape, reasonable stiffness, and acceptable weight. A dash panel may need a clean front face, a predictable thickness at mounting points, and a back face that will not cut wiring or hands. An aerofoil half may need a controlled outer profile and a controlled bonding flange so the upper and lower halves assemble without fighting each other. A honeycomb sandwich panel may need face skins fully bonded to the core, but the exact cosmetic finish of the hidden face may not matter.

That function statement prevents two common mistakes. First, it stops you from building a matched tool for a part that only needs one controlled side. Second, it stops you from using an open mold for a part where the uncontrolled side is not really free. If the back face sets a clearance to a tire, steering column, throttle body, radiator, pedal box, or another laminate, then the back face is part of the design. If the back face is only hidden air, open molding may be enough.

The function statement also catches rule and budget constraints early. Motorsport composites are not chosen in a vacuum. The material and method must fit the technical regulations for the category, and the budget must fit the level of risk. If your class bans carbon in the part you plan to make, the mold choice for a carbon layup is already wrong. If the part is safety critical and you cannot test it properly, a clever tool does not make it acceptable. Tool control is only useful inside the rule and verification envelope you can actually satisfy.

The open-mold case

An open mold is the right default for many home-workshop competition-car parts. You make a pattern, take a female mold from it, prepare the mold surface, lay the laminate into the mold, cure it, release it, trim it, and finish it. The mold gives you the visible surface. Your hands, rollers, fabric choices, and cure management give you the back of the laminate.

The advantage is simplicity. You need one accurate working surface, not two. You can put your effort into making that surface fair, stiff, sealed, and stable. You can see the laminate as you place it. You can chase air out of the gel-coat-adjacent layer. You can add local stiffeners, extra plies, flanges, or joins where the part needs them. If the part is low-volume bodywork, a duct, a cover, a nosecone, or a non-critical panel, this is often the correct trade.

The limitation is that the back face is not geometrically controlled by a hard tool. If you finish with fabric edges, resin-rich ridges, overlaps, or a rough release film texture, that is the back face you get. For many parts that does not matter. For parts that sit near moving hardware, seal to another part, or need repeatable thickness, it matters a lot.

Open mold also depends heavily on tool quality. A cheap, floppy, air-filled mold can sabotage even a simple part. The mold laminate needs enough thickness and stiffness to hold its shape. Small molds for thin lightweight components can be thinner than large body molds, but the principle is the same: if in doubt, err toward a thicker tool rather than a tool that distorts while you are laminating or curing. The layer next to the gel coat matters because trapped air becomes weak spots and can later expand into cracks or blisters if the mold gets hot or sees elevated-temperature cure.

That is why an open mold is not the same as a careless mold. The first reinforcement behind the gel coat may be a fine glass tissue, used to prevent coarse chopped-strand fibers from printing through or disturbing the surface. Internal corners deserve attention because coarse or stiff reinforcement can bridge instead of lying into the corner. If you trap air where the part surface depends on the gel coat, you have built future rework into the tool.

Open mold is also attractive when physical access is already difficult. Large multipart body molds can make one-piece layup awkward enough that you end up with arms, shoulders, or your head inside the mold while laminating. In that case, it may be smarter to mold main areas in stages, leave margins, assemble the mold sections, gel over the joint, and laminate strips over the joint to the same ply count as the main areas. That approach can give slightly less ultimate strength than a fully continuous layup, but for many non-critical components the practical gain in access and workmanship can be worth it.

The matched-mold case

Matched molding adds a second tool face. The laminate is placed in one mold half, and the other mold half is pressed against it by weight, clamps, or a more elaborate press. In industrial use this can involve precision tools, heat, and carefully controlled production. In a home-workshop version, it can be as modest as building a male mold to press the laminate into the original female mold.

The benefit is control. A matched tool can produce a consistent laminate thickness, better consolidation, and a good finish on both faces if that is required. This matters when the back face is visible, when it must fit another part, when thickness affects stiffness or assembly, or when a repeated production run needs less part-to-part variation.

Matched molding is especially useful with carbon fabric in tight curvature. Carbon fabric is stiffer than many other fabric types, and it may not want to conform into a sharp internal radius under hand pressure alone. A male mold pressing on the back of the laminate can push those stiff fibers into areas where they might otherwise lift, bridge, or trap air. That is not only cosmetic. Voids between the first fabric layer and the outer surface, or voids inside the laminate, can leave a part both ugly and structurally deficient.

The penalty is real. You must build the second mold half, release it, maintain it, store it, and make sure it closes correctly. You also lose some visual access during cure. If the laminate shifts, wrinkles, traps excess resin, starves a corner, or bridges under the male half, you may not see it until the part is cured. Matched molding is therefore only sensible when the part requires the extra thickness control, consolidation, two-face finish, or production repeatability.

Shape limits matter. The home-workshop matched-mold method is best for simple, not-too-deep parts whose mold halves can be separated easily. Aerofoil top and bottom halves, cycle-type mudguards, and dash panels are good examples from the corpus because their shapes can justify better consolidation or both-face finish without forcing the tool into a demolding trap. A deep, undercut, complex shape may demand multipart tooling before it demands matched molding, and those are different problems.

Pressure-assisted open molding

There is a useful middle ground between open contact molding and a true matched tool. Flat sandwich panels can be laid up with resin-impregnated plies on each face of a honeycomb sheet, then pressed with a board weighted or clamped in place. Polythene sheet can keep the board from bonding to the laminate. Simple curved sandwich panels can use a conformable weight such as a sandbag to maintain contact during cure.

This is pressure molding in principle, but it is still a legitimate DIY technique for suitable non-critical parts because it uses little specialized equipment. It helps maintain contact between skin, core, and mold without requiring a full male tool. The boundary is important: these improvised pressure methods should not be treated as proof for critical structural components. They are a way to get better quality in non-critical panels, not a substitute for engineering verification.

Vacuum consolidation sits in the same decision family. It adds cost and bagging time, but it can improve quality and mechanical efficiency, especially on intricate parts with deeper shapes and tight internal radii where hand pressure may not keep laminate and mold fully in contact. For simple, essentially flat shapes, contact pressure may be enough. For more intricate shapes, vacuum bagging may be justified before you commit to a matched mold.

Elevated temperature cure is another control layer, not a mold type. A modest post-cure can shorten cure time, allow earlier mold reuse, and improve laminate properties when the resin system benefits from it. But heat also exposes weaknesses in the tool: trapped air near the gel coat can expand, and a marginal mold can distort. If you plan to post-cure, build and choose the tool as if it will see that heat.

The four-question selector

Use four questions to make the choice.

First, which face controls the job? If the outer face defines airflow, appearance, and fit while the inner face is hidden and non-critical, open mold is probably enough. If the inner face sets clearance, seals against another part, becomes a bonding flange, carries hardware, or must look finished, matched molding or a pressure-assisted process moves up the list.

Second, what controls thickness? Hand layup in an open mold can produce useful parts, but it does not inherently give a precise back surface. If the part can tolerate local thickness variation, open mold remains attractive. If thickness changes will move a hinge line, spoil an aerofoil assembly, alter a dash fit, pinch a duct, or change a sandwich panel bond line, add pressure control.

Third, what controls consolidation? Glass chopped strand mat in a forgiving body panel behaves differently from stiff carbon cloth in a tight radius or a honeycomb sandwich skin that must stay in contact with core. If the fabric or core wants to lift away from the mold, the tool must answer that problem. A male mold, board, sandbag, or vacuum bag may be the answer depending on geometry and criticality.

Fourth, how many times will you make the part? A one-off non-critical cover rarely pays back the time to build a true matched tool. A repeated part, or a part whose first failed attempt wastes expensive carbon, honeycomb, resin, and time, may justify more tooling. Do not count only tool cost. Count the cost of rectifying poor surface finish, voids, inconsistent thickness, trimming variation, and scrap.

Sub-skill 1: separate surface accuracy from laminate control

Surface accuracy is the shape of the mold face. Laminate control is what happens through the thickness of the part. You can have a beautiful open mold and still make a poor laminate if the fabric bridges, the corner traps air, the core loses contact, or the back face becomes a resin-rich mess. You can also have a matched tool that controls thickness but reproduces a bad pattern. These are separate skills.

For the mold-choice decision, ask whether the existing problem is surface shape or laminate behavior. If the part needs a fair exterior surface, invest in the pattern and open mold surface. If the part already has a good mold surface but the laminate refuses to stay down, invest in pressure, vacuum, or matched-tool control. If both problems exist, solve surface first, because a second mold half made from a poor first shape only gives you two tool faces that accurately disagree with what the car needs.

This is also where the sibling lesson on deciding when the tool must be more accurate than your eye belongs. That lesson handles the accuracy threshold. This lesson assumes you have a required shape and asks how much of the laminate must be controlled against that shape.

Sub-skill 2: read the demolding geometry

Matched molding fails early if you ignore separation. The corpus is clear that home-workshop matched molds are best for simple shapes that are not too deep and whose halves can be easily separated. That is not a minor convenience. If the male half locks into the part, if the part locks into the female half, or if the closing motion drags fabric out of position, the matched tool becomes a scrap generator.

Read the geometry from the cure state backward. Imagine the resin is hard, the fibers are sharp, and the part is still gripping the tool. Which direction does each half move? Where does the part flex during release? Where can you apply wedges or air without damaging the surface? Does the mold need flanges for clamps? Do those flanges block your hands during layup? Can you trim the part without cutting through carbon edges that destroy blades and leave frayed fibers?

If the answer is uncertain, do not force a matched mold because it sounds more advanced. Use an open mold, a multipart open mold, or a pressure-assisted method until the shape proves it needs the second hard face.

Sub-skill 3: predict fabric obedience

Fabric obedience means whether the reinforcement will stay where you put it until cure. CSM and glass tissue are forgiving around many mold details. Woven carbon is lighter and stiffer in the finished part, but the fabric itself can be less willing to conform. Honeycomb core brings a different issue: the skin must keep full contact with the core, and the core edges may need later sealing to keep the panel weatherproof and presentable.

Your mold choice should reflect the least obedient material in the stack. If you are laying carbon over tight internal corners, matched pressure can prevent the first layer from lifting away from the outer surface. If you are making a flat honeycomb panel, a board and release film may give enough pressure. If you are making a simple curved honeycomb panel, a sandbag may keep contact over the area. If the part is intricate and deeper, vacuum consolidation may be the better pressure source.

The warning is that pressure does not repair a bad layup plan. If the fabric is cut so it cannot sit into the corner, if the ply orientation fights the curvature, or if there is too much material stacked at a radius, the tool will not make the fibers behave. Pressure should consolidate a workable layup, not crush an impossible one.

Sub-skill 4: account for the hidden face

The hidden face is where many tool choices go wrong. In an open mold, the back of the laminate may take the finish of polythene, release film, peel ply, glass tissue, fabric texture, or whatever final layer you used. That may be acceptable. It may even be desirable if the back face will later be bonded, trimmed, or hidden. But if you need a smooth back face, a consistent flange, or a safe hand-contact surface, the hidden face is not free.

Matched molding can give a good finish on both faces, but only if the male mold itself has the finish and release quality you need. A rough male mold gives you a rough inner face. A male mold with poor edge relief can pinch resin and fabric. A male mold without sensible flanges can apply uneven pressure. The second face is a tool, not a magic smoothing board.

For intermediate work, the practical test is simple: if you can describe the allowed hidden-face condition in words and inspect it after cure, open mold may be enough. If you cannot tolerate the hidden face wandering, or if inspection after cure is too late to save the part, choose a method that controls it during cure.

Sub-skill 5: count rework as tooling cost

Open molds feel cheaper because the first tool is cheaper. That is only true if the parts come out usable. If every part needs long finishing work, if the back face must be sanded flat, if voids in tight corners must be repaired, or if carbon appearance is spoiled by bubbles under a clear finish, the cheap tool has moved cost into every part.

Matched molds feel expensive because the second tool is expensive. That is only a bad investment if the second tool controls something the part does not need. If it reduces scrap, controls thickness, improves both-face finish, and makes a production run repeatable, the extra tooling stage can be the cheaper path over the run.

The right accounting includes materials, layup time, cure time, finishing time, defect repair, and mold reuse. A modest warm-box cure can let a mold be reused sooner, which only matters if you are making enough parts for mold turnover to matter. A one-off part may not care. A batch of aerofoil halves might care very much.

Calibration cues

A good open-mold choice feels simple in the shop. You can reach the layup area. The fabric lies down without heroic pressure. The part releases without damaging the mold. The visible surface reflects the mold quality. The back face is exactly as acceptable as you predicted. Trimming exposes no surprising voids in the corners. If you post-cure or leave the mold in sun, the tool does not blister from trapped air.

A bad open-mold choice reveals itself as access trouble, hidden-face trouble, or consolidation trouble. You find yourself trying to reach deep into the mold while breathing too much vapor. You chase bubbles in the same radius repeatedly. The carbon lifts away from the surface. The back face interferes with the car. The part works only after heavy finishing. These are signs that the tool choice was too simple for the control problem.

A good matched-mold choice feels boring in the best way. The mold halves close without forcing. Clamp or weight loads feel even. The released part has predictable thickness. Both faces match their intended finish. Tight radii are consolidated. Repeated parts trim and fit the same way. If the part is an aerofoil half, a dash, or a mudguard, it looks like the tooling controlled the shape instead of the fabric negotiating with it.

A bad matched-mold choice feels like you built a problem twice. Layup becomes slower because the second half blocks your view. Demolding becomes tense. The part has wrinkles you could not see during cure. The mold halves trap the part. The laminate is crushed in one area and loose in another. The second tool face does not improve the part enough to justify the time it took.

Failure modes and recovery

Failure mode one is the vanity matched tool. You build male and female halves because matched molding sounds professional, but the part only needed one finished face. Recovery is to step back to the function statement. If the hidden face is not a fit, safety, clearance, or finish requirement, make the open mold well and spend the saved time on the pattern, mold stiffness, layup access, and release process.

Failure mode two is the heroic open mold. You use one mold face even though stiff carbon fabric, tight curvature, or a core bond line needs pressure. Recovery is to add the smallest control layer that solves the actual problem. That may be CSM strips in tight internal corners before carbon, a glass tissue finish layer, a sandbag, a weighted board, vacuum consolidation, or a true matched male mold. Do not jump straight to the most complicated tool unless the part demands it.

Failure mode three is the trapped-air tool. You build the mold quickly, accept air in the gel-coat-adjacent laminate, and later discover cracks, blisters, or surface defects when the mold warms up. Recovery is prevention, not repair. Use fine tissue behind the gel coat where appropriate, work air out of the first layers, and build enough mold thickness that the tool stays stable.

Failure mode four is the demolding trap. The matched halves make sense while the resin is wet, but not after the part cures. Recovery is to redesign the split, simplify the tool, or add parting sections before you commit to production. If separation is uncertain on paper, make a small trial or choose a less ambitious pressure method.

Failure mode five is treating non-critical pressure tricks as structural proof. A board and clamps over honeycomb, or a sandbag over a curved sandwich panel, can be successful for non-critical items. That does not mean the method is justified for suspension, crash, or high-load structure. Recovery is to keep criticality explicit. If the part must carry serious load, you need design and test evidence, not only a tidy cure.

Failure mode six is ignoring regulations. The part may be beautifully tooled and illegal. Recovery is to check the class rules before buying material or choosing the mold. The regulation question comes before the tooling question because tooling cannot save a banned material.

Choosing between open, pressure-assisted, vacuum, and matched

Use this practical ladder. Start with open contact molding. If the part has one important face, good access, compliant fabric, and no demanding thickness requirement, stay there. Build the mold well. Control the gel coat and first reinforcement. Trim and finish honestly.

Move to pressure-assisted open molding when the part is flat or simply curved, the back face does not need a polished hard-tool finish, but contact through the laminate matters. A weighted board on flat honeycomb or a sandbag on simple curvature can maintain contact without the cost of a male mold. Keep it to non-critical parts unless you have separate proof.

Move to vacuum consolidation when the part is still essentially an open-mold part, but the geometry, internal radii, or laminate quality need better contact than hand pressure can reliably give. Vacuum adds consumables and setup time, but it can save rework on intricate components.

Move to matched molding when the second face itself must be tooled, when thickness control is a requirement, when carbon conformity is a repeat problem, when both-face finish matters, or when production count justifies the second tool. Keep the part simple enough that the halves separate, and remember that a matched tool is a system: female surface, male surface, flanges, pressure method, release, cure, and demolding all have to work together.

What to borrow from professional practice

Professional race-car composite shops separate pattern production, machining, laminating, cure, trimming, and testing because control through the production process matters. You may not have separate rooms, multi-axis machining centers, or CNC-machined polished aluminum tools, but you can borrow the discipline. Keep pattern work distinct from layup decisions. Keep dirty trimming away from clean laminating. Keep the mold decision tied to the part requirement instead of to whichever method seems most advanced.

The corpus includes a pair of CNC-machined aluminum molds made from solid and hand polished, with very high precision and possible mirror finish. That level of tooling is not a normal home-workshop answer. It is useful as a reference point. At one end of the spectrum is a simple open GFRP mold. At the other is a precision matched tool. Your job is to place the race-car part honestly on that spectrum.

The best shop habit is to write down the choice before you start. Record the part function, controlled face or faces, laminate material, pressure method, expected run count, cure plan, and reason for not choosing the next more complicated tool. That record is not paperwork for its own sake. It protects you from tool creep, where every part slowly becomes a matched-mold project, and from false economy, where every part is forced through an open mold even when the laminate is telling you it needs pressure.

Where this lesson stops

This lesson is not the pattern-plan lesson. It does not teach how to turn a part shape into a pattern plan or where to put every flange and split line. It is not the release lesson. It assumes you will apply release correctly without contaminating the layup. It is not the tool-proof lesson. It tells you how to choose the tool concept, then points you to proving the tool before you cure a real part.

The boundary matters because tool choice is an early fork. Once you decide open mold, pressure-assisted open mold, vacuum consolidation, or matched mold, the later lessons become much sharper. Pattern planning answers how to build the shape. Release answers how to get the part back. Proofing answers whether the tool actually behaves. This lesson answers which kind of control the part deserved in the first place.

A final decision rule

If the part has one important face, low criticality, simple access, forgiving fabric, and a small run count, choose the open mold and make it clean, stiff, and reliable.

If the part has a core bond line, simple curvature, and a hidden face that only needs contact rather than cosmetic finish, choose pressure assistance before a full matched tool.

If the part is intricate enough that hand contact is unreliable, but the second face does not need a hard finish, consider vacuum consolidation.

If the part needs both faces, repeatable thickness, carbon conformity in tight radii, or production repeatability, choose matched molding, provided the shape is shallow and simple enough to demold cleanly.

That is the craft: choose the simplest tool that controls the failure mode you can name.

Worked example: low-volume hillclimb nosecone

The Mallock hillclimb and sprint-car nosecone in the corpus is a good open-mold example because the part is bodywork with an exterior shape and local stiffness needs, not a precision two-face molded structural member. The pattern was made from MDF, polyurethane foam block, and body filler, then painted and rubbed down before a GFRP mold was taken. The part itself was made in glass CSM and woven carbon with local stiffeners.

For that kind of part, the controlled surface is the outside of the nose. It has to look right, meet the airflow reasonably, accept the aero package, and fit the car. The hidden face can tolerate more variation as long as it clears the chassis, does not foul hardware, and carries local stiffeners where needed. A full matched mold would add cost and time without automatically improving the part function.

The open-mold discipline still matters. The mold must be stiff enough not to distort the nose shape. The gel-coat-adjacent laminate must be free of air that could later blister. If the nose mold is multipart and access is awkward, staging the layup by main areas and joining inside the assembled mold may be more practical than trying to laminate the whole part in one uncomfortable reach. The final strength of a staged non-critical body component may be slightly lower than a continuous one-piece layup, but the workmanship improvement can be worth it when the alternative is poor access and poor consolidation.

The decision you would write on the job card is: open female GFRP mold, staged layup if access demands it, local stiffeners where the nose needs support, and no matched male tool unless repeat production or hidden-face fit starts causing defects.

Worked example: aerofoil half with carbon in tighter curvature

An aerofoil top or bottom half is where matched molding starts to earn its keep. The corpus names aerofoil halves as parts that can lend themselves to the process, provided the shape is simple enough and not too deep. The reason is not that aerofoils are glamorous. It is that an aerofoil half can care about profile, thickness, bonding edge, inner face, and repeatability all at once.

If you lay carbon into an open female mold for an aerofoil half, the outer face may be accurate while the back face depends on hand compaction. In gentle areas that may be fine. In tighter curvature, stiff carbon can resist sitting into the mold, leaving air between the first layer and the outer surface. That defect is visible if the finish is clear, but it is also a laminate quality problem even when painted.

A male mold pressing the back of the laminate can push the fibers into those tighter radii and hold thickness more consistently. It can also give the inner face a better finish if that face later bonds to the other half or sets an internal passage. The price is a second tool face, a clean closing method, and a demolding plan. If the aerofoil half is too deep or has undercuts, matched molding may become a trap. If it is simple and repeated, the matched tool may reduce rework and make the top and bottom halves assemble with less argument.

The decision you would write on the job card is: matched mold justified if the carbon first ply will not reliably conform under hand pressure, if the inner face or bond flange must be controlled, or if repeated aerofoil halves need consistent thickness. If only the outer cosmetic surface matters and the back face is later trimmed or bonded with generous tolerance, start simpler.

Worked example: flat honeycomb panel without a true matched tool

A flat honeycomb sandwich panel sits between the two extremes. The part may not need a polished inner face, but it does need the skins to stay in full contact with the core during cure. The corpus describes laying resin-impregnated plies on either face of honeycomb sheet, then pressing the laminate mechanically with a board weighted or clamped in place. Polythene sheet can act as release film so the board does not bond to the laminate.

That is not a full matched mold in the precision-tool sense. It is pressure molding in a practical DIY form. The board controls contact, not fine cosmetic geometry. For a non-critical flat panel, that can be the right answer. You get better skin-to-core contact than hand pressure alone without building a dedicated male mold.

For a simple curved sandwich panel, the same logic can use a sandbag as conformable pressure. The bag is not defining a finished inner surface. It is keeping the laminate in contact over the area during cure. The warning is criticality. The corpus explicitly keeps these methods out of critical structural components. If the panel is a cover, duct wall, or non-critical closeout, the method may be sensible. If it carries crash, suspension, or major chassis load, pressure-assisted DIY curing is not proof.

The decision you would write on the job card is: open mold plus board or sandbag pressure for non-critical honeycomb contact, with release film between pressure tool and laminate, and no claim of structural proof without separate testing.

Common mistakes and what good looks like

Mistake one is buying complexity instead of control. You decide on matched molds before naming the defect the second half prevents. Good looks like writing the defect first: uncontrolled back face, poor thickness repeatability, carbon bridging, core contact, two-face finish, or repeated production variation. If none of those are present, a good open mold is probably the better craft.

Mistake two is under-tooling carbon. You treat carbon like forgiving glass in a tight internal radius, then wonder why the clear finish shows bubbles or the corner sounds hollow. Good looks like predicting fabric obedience before layup. Use local glass or CSM support where appropriate, adjust ply cuts, and add pressure, vacuum, or a male mold when stiff fabric will not stay in contact.

Mistake three is making the mold too light. The part looks simple, so the mold is laid up thin and fast. Later it moves, blisters, or prints defects into every part. Good looks like building the mold as a stable tool, not as a disposable shell: enough thickness, careful first layers, minimal trapped air, and a surface that can tolerate the cure plan.

Mistake four is ignoring access. A one-piece layup sounds stronger, but you cannot physically reach the front of the assembled mold well enough to laminate it. Good looks like choosing a layup sequence that produces better workmanship, even if that means molding main areas separately and joining them inside the mold for a non-critical part.

Mistake five is confusing pressure with precision. A board, clamp, sandbag, or vacuum bag can improve contact, but it does not automatically create a precision matched tool. Good looks like naming what the pressure device controls. If it controls core contact, say that. If it controls a finished inner face, it needs to be a real tool face with release, shape, and finish quality.

Mistake six is forgetting the rulebook. A beautifully made carbon part is still wrong if the class bans carbon in that location. Good looks like checking material permission before tool choice and keeping the regulations beside the function statement.

Drill: the four-question mold selector

Run this drill before your next fabrication job. It takes about 45 minutes in the shop and one follow-up review after cure if you make a test coupon.

Choose two real or plausible parts from your car: one bodywork or duct-style part, and one panel, dash, aerofoil half, or sandwich-panel-style part. For each part, answer the four selector questions on paper. Which face controls the job? What controls thickness? What controls consolidation? How many parts will you make?

Then assign one of four methods: open contact mold, pressure-assisted open mold, vacuum-consolidated open mold, or matched mold. You must write one sentence explaining why the next more complicated method is not justified. You must also write one sentence explaining what defect would make you upgrade.

If you have scrap material available, make a small coupon that represents the risky feature: a tight carbon radius, a flat honeycomb skin under a board, or a simple curved area under sandbag pressure. The coupon does not prove the full part, but it teaches whether the material behaves the way your selector assumed. Inspect for trapped air, bridging, poor skin contact, rough hidden face, and demolding trouble.

Success criterion: by the end of the drill, you can defend the mold choice without using the word better. You should be able to say the chosen method controls this specific failure mode, the rejected more complex method would not add useful control for this part, and this visible coupon result either confirms or changes the plan.

When this principle breaks down

The selector breaks down when the part is safety critical and the corpus-supported home-workshop methods do not provide enough design or test evidence. Composite suspension links, crash structures, roll structures, and other high-load parts are not justified by a tidy layup and a clever mold. The testing chunks in the corpus describe tensile, compressive, flexural, shear, peel, tear, repeated-load, environmental, proof, and ultimate-strength testing used to evaluate materials and components. If the part belongs in that world, tool choice is only one input to an engineering verification process.

The selector also breaks down when the shape demands industrial precision. CNC-machined aluminum matched molds can deliver high precision and mirror finish with enough effort, but the corpus treats that as outside the likely reach of most DIY molders. If the part needs that level of precision, the correct answer may be outsourcing, changing the part concept, or choosing a simpler component to build in-house.

Finally, the selector breaks down when regulations decide the answer before workmanship does. If the category disallows the material or type of component, no mold choice can rescue it. The practical sequence is rules, function, controlled faces, consolidation risk, production count, then tool choice.

Author Review

No quiz questions are attached to this lesson.

Sources

#	Document	Chunk	Pages	Score	Collection
1	Competition Car Composites Simon McBeath	c8ea927b-ee2f-add5-6a09-0c2ae6daa1eb	133	1	uio_books_raw_v1
2	Competition Car Composites Simon McBeath	b0b6aa95-bae6-f58f-78aa-3010e487b00a	133	1	uio_books_raw_v1
3	Competition Car Composites Simon McBeath	09d4e626-b413-80f8-1d80-0105e1881ae3	148	1	uio_books_raw_v1
4	Competition Car Composites Simon McBeath	83bc8ccc-3320-7340-25ef-873a72e41eb8	130	1	uio_books_raw_v1
5	Competition Car Composites Simon McBeath	724f6846-659c-ed7f-83d9-8d0189cdf135	85	1	uio_books_raw_v1
6	Competition Car Composites Simon McBeath	29174e7e-f85f-b3ed-98d4-f09ab24ba6b3	105	1	uio_books_raw_v1
7	Competition Car Composites Simon McBeath	b62835e2-37fe-36d0-af44-3b5152d14917	184	1	uio_books_raw_v1
8	Competition Car Composites Simon McBeath	4cd165c8-25b6-009a-f4b5-4fae9a62b8dc	12	1	uio_books_raw_v1
9	Competition Car Composites Simon McBeath	6c01151e-a215-970c-b1db-aa00e94ca228	98	1	uio_books_raw_v1
10	Competition Car Composites Simon McBeath	629cf934-5b41-0aa0-eb70-cec1d94b0bbb	171	1	uio_books_raw_v1
11	Competition Car Composites Simon McBeath	e7681fb9-23a0-7bc7-e029-4ec6bb0c593d	135	1	uio_books_raw_v1
12	Competition Car Composites Simon McBeath	6517b923-9f62-7638-e6e1-0d93afa10f8f	177	1	uio_books_raw_v1
13	Competition Car Composites Simon McBeath	50e8919c-ef19-4354-dea8-95d9c311c69e	178	1	uio_books_raw_v1
14	Competition Car Composites Simon McBeath	a92a57d7-66ad-7c18-c969-cf0c0d4005e9	204	1	uio_books_raw_v1