Medical Cart Shell Rapid Tooling Plastic: How the Process Actually Works
This page explains how a medical cart shell gets from CAD to a molded plastic housing in weeks, not months. It is written for design engineers and sourcing managers who need to judge whether bridge tooling fits their program, and where CNC-machined shells or vacuum casting make more sense.
In this article
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
What Makes a Cart Shell Different From a Normal Enclosure
A medical cart shell is a large, thin-walled plastic housing that carries electronics, drawers, a work surface and casters. It has to stay rigid when a 60 kg pump is rolled over a door threshold. It has to survive repeated wiping with alcohol, quaternary ammonium and diluted bleach without crazing. It also has to look clean in a patient room, which is a cosmetic requirement, not just a structural one.
Those demands pull in opposite directions. Stiffness wants thick walls and glass-filled resin. Chemical resistance wants a specific polymer family. Cosmetic appearance wants a smooth, sink-free surface. One shell can carry all three, but only if the wall, the rib layout and the resin are chosen together rather than one at a time.
Size is the other difference. Cart shells are usually 500–900 mm tall and often over 600 mm in one direction. That pushes past the platen of many small molding presses and past the work envelope of a compact machining center. Tooling and machine selection start with the part envelope, not with the resin.
So the real question is never "which process is best." It is which process holds the tolerances that matter on this shell, at the quantity you can commit to now, without locking your design before clinical feedback arrives.
How Medical Cart Shell Rapid Tooling Plastic Works
Rapid tooling means cutting a production-shaped mold from aluminum or a soft steel insert instead of a hardened multi-cavity tool. The cavity and core are milled on 3-axis and 5-axis CNC centers, then hand-polished. Cycle life lands in the low thousands to tens of thousands of shots depending on resin and wall thickness. That is enough for pilot builds, clinical trials and early market launch.
Aluminum conducts heat about five times faster than tool steel. The mold reaches ejection temperature sooner, so cycle time drops. The same property punishes you if the cooling layout is careless: hot spots appear around thick bosses and cause warpage that no amount of packing pressure will fix.
Gate and runner design matters more on a large thin shell than on a small part. A single edge gate on a 700 mm shell forces the melt to travel too far before it fills, and short shots show up at the far corner. Multiple gates or a hot runner balance the fill, at the cost of visible weld lines that must be placed where the customer will not see them.
Draft is where most first-round tools get cut. A textured ABS shell needs 1.5–3° of draft; a polished PC shell can run 0.5–1°. Undercuts that a CNC-machined prototype tolerated will need side actions or lifters in the mold, which adds cost and adds a maintenance point.
The bridge tool is a bridge in name only if the design changes after trial molding. Aluminum inserts can be welded and re-machined once or twice. After that, a new insert is cheaper than chasing the old one.
Where Rapid Tooling Stops Being the Right Answer
Rapid tooling is a poor fit when the annual volume is already fixed and high. If you know you will ship 80,000 carts a year for five years, a hardened steel tool with optimized cooling pays back long before the aluminum insert wears out. Running aluminum at that volume means replacing inserts mid-program and re-qualifying dimensions.
It is also a poor fit when the shell geometry is still moving. If clinical feedback may relocate a drawer stack or a handle, cut a CNC-machined or vacuum-cast shell first. Those routes change with a new toolpath, not a new mold insert.
Tolerance is the third boundary. Molded shells typically hold ±0.1 mm on molded features, and less on features that shrink differently across a long flow path. If a mating surface must sit at ±0.005 mm, machine that surface after molding or make it a separate machined part.
Chemical resistance sets a fourth boundary. Not every resin survives the disinfectant list a hospital actually uses. A resin that passes one wipe test may craze after two hundred cycles. Ask for the cleaning agent list before the resin is locked, not after the first field complaint.
Design Rules That Decide Whether the Shell Molds Cleanly
Uniform wall thickness does more for shell quality than any other single rule. Aim for 2.5–3.5 mm on a cart shell and keep variation under 25 percent of nominal. A 4 mm boss standing on a 2.5 mm wall shrinks at a different rate and pulls a sink mark onto the visible face.
Ribs add stiffness without adding wall. Keep rib thickness at 50–60 percent of the nominal wall and space them roughly twice the wall thickness apart. Deeper than three times the wall thickness, and the rib becomes its own filling problem.
Bosses need their own treatment. A boss should be tied to a nearby wall with a gusset or a rib rather than standing alone. Boss wall thickness of 60 percent of nominal, with a small core-out, keeps the surrounding surface flat.
Corner radii of at least one wall thickness reduce stress concentration and help the melt turn the corner. Sharp internal corners on a cart shell become crack starters during a caster impact.
Finally, plan the parting line before you plan the cosmetics. A parting line that runs across a visible front panel will be seen, no matter how well the tool is fitted. Move it to a radius or a shadow line early in the layout.
Choosing Plastic for a Cart Shell You Can Disinfect
ABS is the default for cart shells. It machines well, molds well, takes texture and color, and holds cost down. Its weakness is chemical resistance: repeated alcohol wiping dulls the surface over time, and some aggressive disinfectants attack it faster.
ABS/PC blend trades some flow for impact strength and better chemical resistance. It is the common choice for crash carts and anesthesia carts, where the shell takes knocks and is wiped many times a day. It costs more and needs higher melt temperatures, which shortens aluminum tool life a little.
Polycarbonate alone gives the best impact and clarity of the common options, but poor flow in a long thin wall and high sensitivity to moisture before molding. Drying the resin is not optional. Modified PPO is another route: good chemical resistance and dimensional stability, harder to color-match than ABS.
Medical-grade grades cost more than general-purpose grades of the same polymer. If the shell touches the patient or enters a sterile field, that premium is not avoidable. If it is a housing for a monitor arm, a general-purpose grade with a documented cleaning protocol may be enough.
Whatever you pick, run a cleaning compatibility check on actual molded plaques, not on a data sheet.
Running the First Mold Trial Without Losing a Week
A mold trial is not a pass-or-fail event. It is a data-gathering run. Set the press to fill 95 percent of the part on first-stage injection, then transfer to pack. Record the fill pattern, the short-shot location and the cavity pressure. Those three numbers tell you more than a visual inspection of one good part.
Measure the first ten parts after the process stabilizes, not the first ten off the press. Shrinkage settles as the mold reaches steady temperature. Parts 1 through 10 run cold and read small; parts 30 through 40 read true.
Check flatness on the largest panel before you check anything else. Warpage on a 700 mm side panel is the defect that kills cart programs, because it shows up as a gap at the drawer front and cannot be sanded out in production.
Log the process window, not just one setting. Record the range of pack pressures and cooling times that still produce an in-spec part. That window is what the molder will actually run for the next five thousand shots.
If the shell includes a machined interface, a caster plate or a rail, check the molded boss positions against that plate before you sign off. A 0.3 mm mismatch here becomes a rework station later.
What Drives Cost and What the Auditor Will Ask For
Cost in rapid tooling is driven by three things: mold size, the number of side actions, and the polish level. A single-cavity aluminum tool for a 700 mm shell with one side action and a standard texture sits in a very different bracket than a two-cavity tool with four lifters and a high-gloss finish.
Part price is driven by cycle time and scrap rate. Aluminum cools fast, which helps. But a shell that warps on every tenth shot pays for the tool savings in scrap. Fix the cooling layout before negotiating the part price.
On the compliance side, expect questions about material traceability, process records and inspection evidence, not just the resin certificate. For a medical device program, ISO 13485:2016 is the framework most auditors will reference. GreatLight holds ISO 9001:2015, IATF 16949:2016, ISO 13485:2016 and ISO 27001:2022.
Inspection evidence should match the drawing. If the drawing calls out a flatness tolerance on the top panel, the report should show that measurement, not a general dimensional summary.
Keep the change history. An auditor who sees three tool modifications and no record of what changed will ask harder questions than one who sees the same three changes documented.
Machining the Tool and the Metal Parts Around It
The mold insert is only half the job. A cart also carries caster mounting plates, drawer slides, monitor brackets, handles and sometimes a machined frame. Those parts come off the same shop floor and have to fit the molded shell.
GreatLight runs 127 high-precision CNC machines across three wholly-owned plants totaling 7,600 m² in Dongguan, plus a Singapore factory. The fleet includes 16 simultaneous 5-axis machining centers, 12 four-axis mills, 27 three-axis machines and 16 mill-turn centers. Maximum processing size is 4,000 mm, which covers cart frames and long rails.
Standard tolerance is ±0.005 mm, with fine finishes down to Ra 0.2–0.8 μm when a sealing or sliding surface needs it. For a caster plate or a rail, Ra 0.8–1.6 μm is usually the practical target.
Because the shell and its metal interfaces are quoted together, the DFM review catches fit problems before either part is cut. Quotation and free DFM analysis come back within 12 hours, production can start within 24 hours, and parts ship in 3–5 days. There is no minimum order quantity, from one prototype to 10,000+ part runs.
Uploads are handled as confidential, and an NDA is available on request.
Step by Step: From CAD to Molded Cart Shell
- 11. Freeze the interface geometryLock the caster plate, drawer rails and electronics mounting pattern first. These are machined or metal parts and should not move once tooling starts.
- 22. Run a DFM pass on the shellCheck wall uniformity (2.5–3.5 mm), draft (0.5–3° depending on texture), rib thickness (50–60% of wall) and gate location. Expect a written report, not a verbal OK.
- 33. Pick resin and cleaning agents togetherMatch the polymer to the actual disinfectant list. Verify on molded plaques, not data sheets.
- 44. Cut the aluminum cavity and coreMold inserts are milled on 3-axis and 5-axis centers. Tight features and shut-offs are finished to ±0.005 mm before polishing.
- 55. Texture and polish to the cosmetic specConfirm the texture depth first, because it sets the minimum draft. A deep texture on thin draft will drag.
- 66. Trial, measure and set the process windowMeasure parts 30–40 for shrinkage and flatness. Record the pack and cooling range that holds the part in spec.
- 77. Approve first article and release productionSign off on the molded features and the machined interfaces together. Then run the pilot quantity.
Rapid Tooling, CNC Shells and Vacuum Casting Compared
Figures are typical ranges for cart-sized shells; confirm against your own drawing.
| Criterion | Rapid tooling (aluminum) | CNC-machined shell | Vacuum casting |
|---|---|---|---|
| Typical quantity | 300–10,000 shots | 1–50 parts | 10–100 parts |
| Dimensional control | ±0.1 mm on molded features | ±0.005 mm on machined faces | ±0.2 mm, silicone-driven |
| Design change cost | Insert weld or recut | New toolpath only | New master model |
| Surface as delivered | Texture from tool | Ra 0.8–1.6 μm machined | Ra 1.6–3.2 μm from master |
| Resin freedom | Injection grades only | Any machinable plastic | Urethane, limited grades |
| Tooling lead time | Weeks, not months | Days | Days after master |
| Best use | Pilot and launch builds | Design still moving | Small clinical batches |
The Short Version
If your design is still moving or you need parts this month, cut a CNC-machined or vacuum-cast shell and leave the mold alone. If the shell geometry is frozen and you need 300 to 10,000 units in a production resin, cut an aluminum bridge tool and start the pilot run.
Frequently Asked Questions
How long does an aluminum bridge tool last?
It depends on resin and wall thickness. General-purpose ABS in a 3 mm wall typically gives several thousand to tens of thousands of shots before wear shows on shut-offs and gate areas.
Glass-filled resin and high melt temperatures shorten that. If the program looks like it will exceed the insert life, plan a steel replacement rather than running the aluminum tool to failure.
Can we change the design after the mold is cut?
Minor changes are possible. Aluminum inserts can be welded and re-machined once or twice, and small steel inserts can be added at a shut-off.
A change that moves a gate, a parting line or a large rib network usually means a new insert. That is why interface geometry should be frozen before the cut starts.
Why does a molded shell hold looser tolerance than a machined part?
Plastic shrinks as it cools, and shrinkage varies with wall thickness, flow length and pack pressure. The same cavity can produce parts that differ by a few tenths of a millimeter across a long flow path.
Machined plastic is cut to size after the material is stable, so it holds ±0.005 mm. If a feature needs that, machine it after molding or make it a separate part.
Which resin should we use for a cart that gets wiped down daily?
Start with the actual cleaning agent list. ABS is the cost default and handles mild alcohol wiping. ABS/PC blend and modified PPO hold up better against repeated aggressive disinfectants.
Verify on molded plaques rather than relying on a data sheet, because dwell time and concentration matter as much as the chemistry.
Do we need a two-cavity tool?
Only if volume justifies it. A two-cavity tool roughly doubles mold cost and increases press tonnage, and it makes fill balance harder on a large thin shell.
For pilot and launch quantities, a single-cavity aluminum tool is usually the better first step. Add cavities when the annual volume is known and stable.
What should be in the first article inspection report?
The measurements the drawing actually calls out, taken on parts from a stabilized process. Include flatness on large panels and the position of any molded boss that mates with a machined part.
Add the process window that produced those parts, so the molder can reproduce them without guesswork.
Send Us the Shell and the Metal Parts Around It
Upload your CAD and drawing set. You get a quotation and a free DFM analysis within 12 hours, covering the molded shell, its caster plates and its rails in one review.
12-hour quote100% inspectionISO 13485:2016No minimum order quantity
More From GreatLight
We publish setup notes, tooling trials and inspection data from the factory floor.
