Mold parts aren't that mysterious once you've torn down a few tools for repair. The core makes the inside surface, cavity makes the outside, and everything else is just there to support those two pieces or move stuff around. But the details get messy depending on what you're molding and how long the tool needs to last.
P20 steel is what most domestic shops use for general purpose tooling. Runs about $5.40/lb from Castle Metals when we ordered 1800 pounds in March 2024. That's pre-hardened to 30 HRC so you can machine it without heat treat, which saves weeks off the schedule. For a medium-sized mold base maybe 18"x24", you're looking at 400-500 pounds of steel just for the core and cavity blocks.
H13 costs more and machines slower but you get better wear resistance. We only use it when the customer is running glass-filled materials or wants 200k+ cycle life. The hardness after heat treat is around 50 HRC versus 30 for P20, which means carbide tooling and slower feeds. EDM becomes necessary for any fine features because you can't mill hardened steel efficiently.
Ejector pins seem simple until they're not
Standard round ejector pins from Progressive Components cost maybe $8-12 each depending on diameter. The problem isn't cost, it's getting them to work reliably for tens of thousands of cycles without galling in the bore.
We did a tool last year for automotive interior trim - 32 ejector pins because the part had complex geometry. After 8,000 cycles, six pins were seizing intermittently. Tore the mold down and found galling on the pin shafts where they ran through the ejector plate. The bore finish was 32 Ra when it should've been 16 Ra or better. That's on us for not catching it during mold acceptance.
Fixed it by pulling all the pins, polishing the bores to 12 Ra, and installing nitrided pins. Nitrided pins cost about $14 each versus $9 for standard, but the surface hardness jumps to 65 HRC. Tool ran another 40,000 cycles before we saw any issues.
Blade ejectors are different - you're wire EDM'ing a custom shape instead of using round stock. Takes forever and costs accordingly. Made a set for a battery housing that ran $2,100 in machining time when round pins would've been maybe $400 total. But round pins wouldn't have worked with the deep ribs in that part.
Cooling channels matter way more than people think
Cycle time is basically controlled by how fast you can cool the part. We measured this on a polypropylene housing - 8mm wall thickness, 180°F mold temp. With 3/8" diameter cooling lines spaced 1.2" from the cavity surface, cooling time was 38 seconds. Moved the lines to 0.7" spacing and got it down to 26 seconds.
That 12 second reduction doesn't sound like much until you're running 500,000 parts per year. Saves you about 1,660 hours of press time annually, which at $65/hour machine rate is over $100k in reduced manufacturing cost. Yet the cooling line redesign only added maybe $3,500 to tool cost.
Conformal cooling using 3D printed mold inserts can cut another 20-30% off cooling time, but you're paying $95-115 per cubic inch of printed material. We did cost analysis on this for a medical device mold - the printed insert was going to be $18,000 versus $4,000 for conventional drilled cooling. Payback was 14 months at their production volume so they passed on it.
Temperature control units are another thing. Cheap ones from China cost $1,800-2,500 and have terrible temperature stability - we measured ±8°F variation on one unit. Thermal Care units run $6,500-9,000 but hold ±1°F. For engineering resins where mold temp precision matters, that's worth paying for.
Flow rate through cooling channels needs to hit turbulent flow - Reynolds number above 4000 or so. We tested this with different pumps on a tool - at 2 GPM through 3/8" lines, flow was laminar and cooling sucked. Bumped it to 5 GPM and temperatures evened out. Going to 8 GPM didn't help much more, just wasted pump capacity.
Gate types and the mess they cause
Edge gates leave a visible mark that needs trimming. Simple to machine but you're paying someone to manually trim parts or setting up a trimming fixture. For industrial parts nobody cares, for consumer products it's unacceptable.
Submarine gates shear off during ejection so there's no manual trimming. The gate channel tunnels under the part surface at 5-7 degree draft. When draft angle is wrong or the gate diameter is oversized, the gate doesn't shear cleanly and you get torn material at the gate location.
Had this happen on a cosmetic housing part - gate was specified at 0.060" diameter but got machined at 0.072" because someone misread the drawing. Parts came out with rough gate scars that didn't meet appearance specs. Took us a full day to rework the gate area and test shots before we were back in spec.
Hot runners eliminate runner waste entirely but you're spending $18,000-30,000 on a manifold system for a 4-cavity tool. Maintenance is constant - nozzles clog, thermocouples fail, heater bands burn out. We budget $1,500-2,500 per year for hot runner repairs on moderately complex tools.
Some materials won't run in hot runners at all. We tried running PVC in a hot runner once - terrible idea. The material degrades at elevated temperatures and you get HCl offgassing that corrodes everything. Ended up converting back to cold runner despite the material waste.
Valve gates give you precise gate control but each valve pin assembly costs $1,800-2,400. For an 8-cavity tool that's nearly $20,000 just in gate hardware, plus you need hydraulic or pneumatic actuation. Only makes sense for high-end applications where gate cosmetics and shot-to-shot consistency are critical.
Side actions fail in predictable patterns
Angle pins pull slides back as the mold opens. Standard angle is 18 degrees from vertical - steeper than that and you get excessive side loads, shallower and the pin needs to be too long. We use 15-20 degree range depending on space constraints.
The pins fail where diameter changes - like going from 0.500" to 0.375" diameter. Sharp transition radius creates stress concentration and you get fatigue cracking after 25k-40k cycles. Had a batch of angle pins fail at 32,000 cycles on a glass-filled nylon tool. Changed the transition radius from 0.020" to 0.080" and nitrided the pins, which solved it.
Wear plates under the slides need replacement every 60,000 cycles or so. Bronze-filled PTFE is standard material - costs about $45-65 per plate depending on size from Misumi. Takes maybe 90 minutes to replace if you've designed for serviceability. Some tools have the wear plates buried where you need 4 hours of teardown to access them. Bad design.
Hydraulic core pulls skip the angle pin entirely - hydraulic cylinder just pushes the core in and out. Way more flexibility in actuation direction and you can use larger cores. The cylinder costs $1,400-2,200 depending on bore size and stroke length. Most molding machines have auxiliary hydraulics that can run 2-3 cores, beyond that you need external power unit.
Venting gets ignored until you have burn marks
Trapped air compresses during injection and heats up to 700-800°F, which burns the plastic and leaves black marks. Standard vents are 0.0008-0.0012" deep at the parting line - deep enough for air but too shallow for plastic to flash through.
In practice, vents clog up with mold release and degraded resin. You're cleaning them weekly with brass brushes or bead blasting. Takes about 30 minutes per mold if vents are accessible, longer if they're buried in some complicated shutoff area.
Porous vent plugs cost $95-140 each but work way better than machined vents in tight spots. The sintered metal is porous at 25-30 micron scale so gas escapes but plastic can't penetrate. We put these in deep rib areas or complex geometry where conventional venting doesn't work.
Some builders are using laser ablation for micro-venting now - burns channels 0.0003-0.0006" deep that are less prone to flashing. Haven't tried this ourselves because we don't have laser equipment. Probably need to be doing aerospace or medical work to justify that investment.
When cheap components destroy your day
Bought discount ejector return pins one time to save $180 on a tool build. The pins measured 36 HRC instead of the specified 48 HRC minimum. After 4,200 cycles the pin heads mushroomed and jammed in the ejector plate. Mold locked up completely.
Had to tear everything down, replace all six return pins with proper hardness units, and clean up the damaged ejector plate holes. Lost three days of production and burned $1,800 in emergency repair labor. That's what happens when you cheap out on $30 worth of components.
Leader pins and bushings control mold alignment. Standard clearance is 0.0007-0.0010" diametral for smooth operation without slop. We spec diamond-coated leader pins from DME at $135 each for tools running abrasive materials - they last maybe 6-8 times longer than uncoated pins.
Sprue bushings from Strack or Hasco cost $80-120 each depending on nose radius and gate diameter. The radius where the sprue meets the nozzle matters - too sharp and you get stress concentration and eventual cracking. We use 0.500" radius minimum on cold runner tools, larger for hot runner manifolds.
Insert molding gets complicated
Metal inserts need precise locating - even 0.015" misalignment can cause dimensional problems. Spring-loaded locating pins hold the insert against datum surfaces during mold closing. The pins cost maybe $25-40 each from Misumi or Fibro, pretty standard component.
Retention of the insert depends on mechanical interlocking. Knurling on cylindrical brass inserts improves pullout force by 35-40% compared to smooth inserts based on testing we did in nylon 6/6. Diamond knurl works better than straight knurl because you get more surface area.
We did an overmold project where ABS first shot needed 5 hours of atmospheric exposure before TPE second shot. Something about dimension stabilization before overmolding - the engineer didn't explain it great but that's what the resin supplier recommended. Meant we needed buffer inventory and rack space, which complicated production scheduling.
Two-shot molds avoid the inventory problem by doing both shots in one cycle. The mold rotates or indexes between stations. But tooling cost jumps to $45,000-65,000 versus $12,000-18,000 for sequential molding. Only makes sense at really high volumes.
Why aluminum tooling still has a place
Aluminum machines 4x faster than tool steel. A cavity that takes 28 hours to mill in H13 takes maybe 7 hours in QC-10 aluminum. At $90/hour shop rate that's serious money.
Tool life is the issue - aluminum is good for 8,000-12,000 cycles depending on resin and conditions. Fine for prototyping or bridge tooling but you can't run production volumes. QC-10 aluminum costs about $9.50/lb versus $5.40 for P20 steel, so material cost is higher but machining savings more than offset it.
Some builders use aluminum cavities with steel cores and gate inserts. We did this for a 35,000 part run - saved about $6,500 in tooling cost versus all-steel construction. Aluminum cavity wore out around 32,000 cycles and needed minor polishing but made it through the run.
Kirksite is even cheaper than aluminum but I've never seen it used for actual injection molding. Maybe for low-pressure applications but not standard injection.
Suggested Addition for Article End
If you're sourcing Injection Molding Parts and want to avoid the headaches we've covered here, find a shop that actually understands tooling fundamentals. Cheap quotes mean nothing when the mold fails at 15,000 cycles because someone used garbage components or skipped proper venting.