Daniel sent us this one — he's deep in the new apartment setup, got the industrial shelving, the Euro boxes, and he's staring at toilet rolls and kitchen paper thinking there's got to be a way to make these fit perfectly. The hospitality industry has injection-molded HDPE inserts for wine bottles and soft drink cans that slot right into a sixty by forty box. But nobody makes one for the specific diameter of his kitchen roll. His question is: can he design this himself in CAD, and if he wants a short run of a few dozen units in HDPE, is there actually a freelance ecosystem that can produce that?
This is exactly the kind of problem that sits in a manufacturing blind spot. You've got standardized containers — the Euro box, sixty by forty centimeters, internal usable area about five sixty by three sixty — and a whole catalog of off-the-shelf inserts for hospitality. But the moment your use case is "I want to store toilet paper in deep storage," you fall off the edge of what the injection molding industry considers worth tooling up for.
Which is weird when you think about it. Every household uses toilet paper. Every household has storage problems. The overlap seems obvious.
It does, but the economics of injection molding punish niche applications. A single-cavity mold for a sixty by forty centimeter part runs five to fifteen thousand dollars. That's steel tooling, the kind that lasts for hundreds of thousands of cycles. Nobody's amortizing that across a run of twenty-four inserts for one apartment. So the market simply doesn't make the product.
We're in this strange gap where the problem is universal but the solution doesn't exist at consumer scale. Which means Daniel's question is really: what's the cheapest way to manufacture something that doesn't exist yet, in quantities too small for traditional manufacturing, in a material that's surprisingly difficult to work with?
Let's start with the design, because that's the part people assume is the barrier. It's not. Fusion 360 — free for personal use if your annual income from design work is under a thousand dollars — has a rectangular pattern tool that can generate a grid in under thirty minutes. You define one cell, specify the spacing, and it replicates across the surface. A beginner following a YouTube tutorial can produce a parametric grid insert in an afternoon.
What does the actual design look like? Daniel mentioned an ideal configuration.
If you're working with a sixty by forty box, you've got about five hundred sixty by three hundred sixty millimeters of usable space internally. Standard kitchen paper rolls are roughly twelve centimeters in diameter. Standard toilet rolls are about eleven. So a practical layout might be six cells for kitchen paper — two rows of three — and twelve cells for toilet rolls, arranged in a grid below or beside them. Each cell needs to be slightly oversized to account for brand variation. Toilet roll diameters vary by brand by up to two centimeters. Design for the largest you use, then shim with foam tape for the smaller ones.
The design phase isn't the bottleneck. It's what happens next.
And this is where Daniel's instinct was correct — the moment he thought "I could 3D print this" and then immediately thought "that makes zero sense," he was onto something. But let me unpack exactly why.
The material he wants is HDPE. High-density polyethylene. It's what those hospitality inserts are made of. It's food-safe, chemical-resistant, doesn't absorb moisture, and it's tough without being brittle. Perfect for storage. The problem is, HDPE is a nightmare to print on a consumer FDM printer.
How much of a nightmare are we talking?
Three specific failure modes. HDPE shrinks one and a half to two and a half percent linearly as it cools. On a sixty-centimeter part, that's up to fifteen millimeters of dimensional change. The part literally pulls itself apart as it prints. Second, bed adhesion. HDPE doesn't stick well to standard print surfaces. You need a heated bed at a hundred to a hundred twenty degrees Celsius and often a specialized adhesive. Large flat surfaces — exactly what a grid insert is — are the worst-case geometry for warping. The corners lift, the print fails, and you've wasted twelve hours.
The material that's ideal for the finished product is the worst possible choice for the manufacturing process.
That's the cruel irony. You can print in PETG — much easier, good chemical resistance, prints at two hundred thirty to two hundred fifty degrees on a standard bed at seventy to eighty degrees. But PETG isn't HDPE. It's more brittle over time, especially in thin walls. PLA is even easier but has almost no heat tolerance and gets brittle with age. For something you're going to be sliding toilet rolls in and out of for years, you want that HDPE toughness.
Let's say Daniel ignores the material problem and goes with PETG. What does a run of twenty-four inserts actually cost in the real world?
This is where the economics get sobering. Each insert for a sixty by forty box, with grid walls maybe two millimeters thick and five millimeters of spacing between cells, is going to use about three hundred grams of filament. At current PETG prices, that's roughly fifteen to twenty-five dollars in material per insert. A Prusa MK4 would take twelve to eighteen hours per insert. And you're going to lose ten to twenty percent of prints to failures on complex geometries with large flat surfaces. So to get twenty-four good inserts, you're printing maybe twenty-eight to thirty. That's five hundred to seven hundred dollars in material and two to three weeks of continuous printing — assuming you own the printer and your time is free.
Which it isn't.
Which it absolutely isn't. And that's before we even discuss the electricity cost or the fact that your printer is tied up for three weeks straight. If you don't own a printer and you're using a service, the per-unit cost goes higher because they're charging for machine time and labor.
3D printing a batch of these is slow, expensive, and prone to failure. Daniel mentioned there's a freelance world for both design and production — let's talk about that.
The freelance ecosystem for this is surprisingly mature. For the CAD design, platforms like Fiverr and Upwork have mechanical designers who will produce a parametric grid model for fifty to a hundred fifty dollars. You send them the dimensions — box internal measurements, roll diameters, wall thickness preference — and they send back a STEP file ready for manufacturing. Treatstock aggregates over fifteen hundred 3D printing services globally, with pricing starting around ten cents per cubic centimeter for standard PLA.
We've established PLA is the wrong material and PETG is questionable. What about getting HDPE printed?
That's where you need to move upmarket. Consumer 3D printing services mostly run PLA, PETG, and maybe ABS or nylon. HDPE requires an industrial machine with an actively heated chamber — something like the re:3D Gigabot X, which maintains a chamber temperature of sixty to seventy degrees Celsius and can handle HDPE's warping tendencies. Services like Xometry, Craftcloud, or JLCPCB's 3D printing division can do this, but they're more likely to default to nylon PA12 via SLS — selective laser sintering — which is a powder-bed process that doesn't have the warping problem at all. PA12 is actually a great material for this application: tough, chemical-resistant, dimensionally stable. But it's not HDPE, and it costs more.
If you search for "3D printing service" and upload an STL file, you're probably getting nylon, not HDPE, and you're paying a premium for it.
And this is the point where most people either give up or massively overpay. But there's a third path that almost nobody thinks of.
Don't print it at all. Cut it from sheet stock.
HDPE sheet is a commodity product. You can buy a six-millimeter-thick sheet, sixty by forty centimeters, for maybe ten to fifteen dollars. It's the same material as the injection-molded inserts, just in flat form. Now, instead of building up the grid layer by layer in a 3D printer, you cut the grid pattern out of the sheet. The cells where the rolls sit become cutouts. The remaining material is your grid walls.
It's subtractive instead of additive. You're removing material rather than depositing it.
And there are two main ways to do this. CNC routing uses a rotating cutting bit to trace the pattern. SendCutSend charges roughly fifteen to thirty cents per square inch for HDPE routing, plus a setup fee of fifteen to twenty-five dollars. For a sixty by forty part with a grid of eighteen cells, you're looking at twenty to forty dollars per insert, no tooling fee beyond the initial setup. Twenty-four inserts at thirty dollars each is seven hundred twenty dollars. Delivered in one to two weeks.
That's competitive with the PETG 3D printing cost, but you get actual HDPE.
It's faster. No twelve-hour print times. The CNC machine cuts the grid in minutes per part. The trade-off is that CNC routing leaves square inside corners — the cutting bit is round, so internal corners have a radius equal to the bit diameter. For a grid insert, that's cosmetic. The rolls don't care if the corners are slightly rounded.
You mentioned two ways to cut sheet stock. What's the second?
This is the dark horse. A waterjet uses a high-pressure stream of water mixed with abrasive garnet to cut through material. It leaves a clean edge, no heat-affected zone, and it can cut HDPE like butter. The pricing is typically ten to twenty cents per linear inch of cut path. For a grid with eighteen cells, each cell requiring maybe fifteen inches of cut perimeter, you're looking at roughly two hundred seventy linear inches of cutting per insert. That's twenty-seven to fifty-four dollars per insert in cutting cost alone — plus the sheet material. So per insert, maybe thirty-five to sixty dollars total.
That doesn't sound cheaper than CNC.
It's not, for a single-piece grid. But here's where it gets clever. Instead of cutting the grid as one piece, you waterjet individual strips with interlocking slots — like the cardboard dividers in a wine box, but in HDPE. Each strip has half-depth slots at the intersection points. You slide them together, and they form a grid. The cutting path is much simpler — just straight lines with notches — so the linear inches drop dramatically. Per-insert cost falls to five to ten dollars for the cut pieces, plus manual assembly. Glue the joints with a solvent adhesive made for polyethylene, or design them as snap-fits, and you've got a rigid grid for under fifteen dollars per insert.
For twenty-four inserts, we're talking a hundred twenty to two hundred forty dollars total, versus seven hundred for CNC or five hundred to seven hundred for 3D printing in PETG.
It's HDPE. The actual material you wanted in the first place. This is the counterintuitive winner for low-volume production of grid-based geometries. Nobody thinks of it because we've been trained to assume 3D printing is the solution to custom manufacturing. But for flat grids, cutting sheet stock is two to three times cheaper and ten times faster.
What's the catch?
You're gluing or snapping together a grid from individual strips. It won't look as seamless as a single molded piece. The joints are potential failure points over years of use, though a good solvent weld on HDPE is essentially a chemical bond — the material fuses at the molecular level. And you need to design the interlocking geometry correctly so the grid doesn't rack or twist.
Which brings us back to the CAD question. Is designing interlocking strips harder than designing a monolithic grid?
You're modeling individual strips with notches instead of one big grid. The notches need to be the width of the material — six millimeters — and half the depth of the strip. Fusion 360 can handle this. A freelancer on Fiverr could definitely handle this. It's maybe an extra hour of design work compared to the monolithic grid.
Let me pull back and ask the business question Daniel raised. What kind of company actually does this? If he searches for "3D printing service" or "custom plastic fabrication," what should he be looking for?
For the waterjet approach, search for "waterjet cutting service" or "CNC plastic fabrication." Look for shops that specifically mention HDPE or polyethylene in their materials list. SendCutSend does laser cutting but not waterjet — they're primarily a metal shop that also does some plastics. For waterjet, you want a dedicated waterjet service like Big Blue Saw or a local fabrication shop. Many cities have job shops with waterjet machines that will take small orders. The key search terms are "custom plastic fabrication," "CNC routing HDPE," or "waterjet cutting service near me.
If he wants to go the 3D printing route despite everything we've said?
Then he wants an industrial 3D printing service, not a consumer one. Search for "industrial 3D printing HDPE" or "large format 3D printing service." Look for companies that mention heated chambers and specifically list HDPE as a material. Treatstock lets you filter by material, so you can search for providers that offer HDPE. But expect to pay a premium — HDPE printing on industrial machines runs maybe fifty to eighty cents per cubic centimeter, which for a three-hundred-gram part is a hundred fifty to two hundred forty dollars per insert. That's five to six thousand dollars for twenty-four inserts. At that price, you might as well pay for the injection mold.
Which is the point where most people's eyes glaze over and they just stack the toilet rolls loose in the box.
Honestly, for deep storage, loose stacking works fine. The insert is an optimization — it keeps rolls from shifting during transport, it makes inventory counting trivial, and it satisfies a very specific aesthetic urge. Daniel's not wrong to want it. But the manufacturing question is genuinely interesting because it exposes how much of our intuition about "just 3D print it" breaks down at scale.
Let's talk about that scale question. At what point does injection molding actually make sense?
If you're making a hundred or more inserts, rapid injection molding with aluminum tooling starts to become viable. Protolabs, RapidDirect, Star Rapid — these companies offer aluminum molds that cost fifteen hundred to three thousand dollars for a part this size, with per-unit costs of fifteen to thirty dollars. At a hundred units, you're looking at three thousand to six thousand dollars total, or thirty to sixty dollars per insert. That's still more expensive than waterjet assembly, but you get a seamless, professional part. At five hundred units, the per-unit cost drops to maybe eight to twelve dollars, and you're in the range where it's cheaper than any other method.
The crossover point is somewhere between a hundred and five hundred units. Below that, cutting sheet stock wins. Above that, injection molding takes over. 3D printing never wins at any scale for this geometry.
Not for grid-based flat parts, no. 3D printing wins when you have complex three-dimensional geometries that can't be cut from sheet — organic shapes, internal channels, lattice structures. But a grid is fundamentally two-dimensional. It's a flat pattern extruded in the Z-axis. Cutting it from sheet stock is the process it was born for.
There's a broader point here about the manufacturing landscape. We've spent a decade hearing that 3D printing will democratize production, and it has for certain things — prototypes, figurines, brackets, one-off replacement parts. But the moment you need more than a handful of units, traditional subtractive methods often win on cost, speed, and material properties.
The waterjet-plus-assembly approach is a perfect example of hybrid manufacturing that the "3D printing will solve everything" narrative misses entirely. You're using a digital fabrication method — waterjet cutting from a CAD file — but you're assembling the final part manually. It's not fully automated, but it's fast, cheap, and uses the exact material you want. This is how small-batch manufacturing actually works in practice. You find the process that matches the geometry and the material, not the process that's trendy.
Let's give Daniel the actionable plan. He wants twenty-four inserts for sixty by forty Euro boxes, configured for his specific toilet roll and kitchen paper diameters. What does he do this weekend?
Step one: measure. Get calipers or a ruler and measure the actual diameter of the largest toilet roll and kitchen paper roll you buy. Add three to five millimeters of clearance per cell. Write those numbers down.
Step two: decide whether to design it yourself or hire it out.
If you've never touched CAD, go to Fiverr. Search for "mechanical CAD design" or "Fusion 360 designer." Send them your box internal dimensions — five hundred sixty by three hundred sixty millimeters — your cell diameters, your preferred wall thickness of two to three millimeters, and tell them you want either a monolithic grid or interlocking strips for waterjet cutting. You'll get a STEP file back in two to three days for fifty to a hundred fifty dollars.
Step three: choose your manufacturing method. What's your recommendation?
For twenty-four units, waterjet-cut interlocking strips. Find a waterjet service — search "waterjet cutting HDPE" or use a platform like Treatstock to find a local shop. Send them the CAD files for the individual strips, specify six-millimeter HDPE sheet, and ask for a quote. Expect to pay five to ten dollars per insert for the cutting, plus material. Assemble with a polyethylene-compatible adhesive — something like Loctite Plastics Bonding System, which includes a surface activator that makes HDPE bondable.
Design: fifty to a hundred fifty dollars. Cutting and material for twenty-four inserts: a hundred twenty to two hundred forty dollars. Adhesive: fifteen dollars. Total: roughly two hundred to four hundred dollars, delivered in one to two weeks. Compare that to five hundred to seven hundred dollars for 3D printing in PETG, or five thousand plus for industrial HDPE printing, or fifteen hundred to three thousand in setup fees alone for injection molding.
That's a compelling number. And if he wants to go even cheaper and more hands-on?
Buy a sheet of six-millimeter HDPE — sixty by forty centimeters — from a plastics supplier. That's ten to fifteen dollars. Get a jigsaw with a fine-tooth blade designed for plastics. Mark your grid pattern on the sheet with a pencil and a straightedge. Cut the strips, cut the notches with a coping saw or a router table if you have one, and assemble. You'll spend a weekend and about a hundred dollars in materials and tools. The result won't be as precise as waterjet cutting, but it'll work, and you'll have the satisfaction of having built it yourself.
There's something appealing about that. The jigsaw approach has a certain dignity.
It's the same impulse that drives people to build their own furniture. You're not doing it because it's cheaper than IKEA — you're doing it because the thing you build fits exactly and you made it.
Before we wrap up the practical advice, there's one more misconception I want to flag. We've been talking about HDPE as the ideal material, but is it actually necessary for this application? Toilet paper isn't corrosive. The insert isn't going outdoors. Does Daniel really need HDPE, or is that just what the hospitality inserts happen to be made of?
That's a fair question. For indoor storage of paper products, PETG would work fine. It's not going to degrade from contact with toilet paper. The reason HDPE is used in hospitality is that those inserts get washed — sometimes in commercial dishwashers — and they need to survive years of abuse. For a home storage insert that sits on a shelf and holds paper rolls, the material requirements are much lower. You could print in PETG and it would last for decades.
The material purism might be overthinking it.
It might be. But the cost argument still favors cutting sheet stock over 3D printing, regardless of material. Even if you used PETG sheet — which exists, though it's less common than HDPE — the CNC or waterjet approach would still be faster and cheaper than FDM printing for a batch of twenty-four.
Which reinforces the core insight: for flat grid geometries, subtractive beats additive at any quantity above one or two units.
That's the thing I want listeners to take away. The right manufacturing process depends on the geometry, not on what's new or exciting. A grid is a two-dimensional pattern. Cutting it from sheet stock is the obvious answer once you stop thinking of 3D printing as the default.
There's one more angle I want to explore before we move to takeaways. Daniel mentioned the freelance ecosystem — designers on Fiverr, printing services on Treatstock. What's the actual experience of using these platforms for a project like this? Is it smooth, or are there pitfalls?
The CAD design part is straightforward. Mechanical designers on Fiverr and Upwork do this kind of work routinely. The key is to be specific in your requirements: provide exact dimensions, specify the file format you need — STEP or IGES for manufacturing, not just STL — and ask for a screenshot of the design before they deliver the final file. Most designers will do a revision or two as part of the base price.
The manufacturing side?
When you upload a file to a service like Treatstock or Craftcloud, you're getting quotes from multiple providers. The prices can vary by a factor of three for the same part. Some providers will flag issues with the design — walls too thin, features too small for their process — and others will just print it and send you a failed part. Read the reviews. Look for providers that have done similar work. And if you're doing waterjet cutting, call the shop and talk to a human. Waterjet shops are often small businesses where the owner answers the phone, and they can give you practical feedback on your design that an automated quoting system won't.
The platform aggregates options, but the human touch still matters.
Especially for an unusual request. A waterjet shop that normally cuts steel brackets for construction equipment might look at your HDPE grid strips and think "huh, that's different" — but they'll quote it, cut it, and ship it. The key is to not be intimidated. These shops want work, even small jobs.
Let's consolidate this into a clear recommendation. If someone listening has the same problem Daniel has — a Euro box, a specific item they want to store in a grid, and no off-the-shelf insert exists — what's their path?
Design for sheet cutting, not 3D printing. Whether you do the CAD yourself in Fusion 360 or hire it out, design the part as either a flat grid with cutouts for CNC routing or as interlocking strips for waterjet cutting. Choose the material based on your actual requirements — HDPE if you need toughness and washability, PETG or even plywood if you don't. Get quotes from at least three services. And don't assume the answer is additive manufacturing just because that's what you've heard of.
The cost range for a single sixty by forty insert, done this way?
Five to forty dollars per insert, depending on method and quantity. At twenty-four units, you're probably around ten to fifteen dollars each for waterjet strips plus assembly, or twenty-five to thirty-five each for CNC-routed single-piece grids. Both are cheaper than 3D printing and both give you the material you actually want.
That's a satisfying answer to a very specific problem. Before we close out, I want to touch on the bigger trend here. We're seeing on-demand manufacturing services — Xometry, Protolabs, SendCutSend — get better at instant quoting, faster at turnaround. Is there a future where custom Euro box inserts are as easy to order as a pizza?
We're getting closer. Xometry's instant quoting engine can now handle a surprising range of processes — CNC machining, sheet cutting, 3D printing in multiple materials — and give you a price in seconds. The barrier isn't the technology anymore; it's the design literacy. Most people don't know how to specify a part in a way that a manufacturing service can quote. They don't know what a STEP file is, or what wall thickness is practical for a given process. That's the gap that freelance designers fill, and it's why the Fiverr-plus-manufacturing-service pipeline is so powerful. The designer translates your idea into manufacturable geometry, and the service produces it.
The bottleneck is human knowledge, not machine capability.
As it often is. The machines are ready. Most people just don't speak their language yet.
Which is probably a good place to land. Daniel's question opened up a whole landscape of small-batch manufacturing that most people don't know exists. The answer isn't "3D print it" — it's "design for the process that matches your geometry, and there's probably a service that can do it for less than you think.
If all else fails, buy a sheet of plastic and a jigsaw. Some problems are best solved with sweat and swearing.
Now: Hilbert's daily fun fact.
Hilbert: In 77 CE, Pliny the Elder recorded transient lunar phenomena described as "a bright point of light" on the moon's surface, which modern analysis suggests may have been the outgassing of sodium-rich vapors from a subsurface deposit — a chemical signature consistent with certain terrestrial feldspar compositions.
...right.
Here's the open question we'll leave you with. As 3D printing speeds keep climbing — Bambu Lab's latest machines are pushing two hundred millimeters per second — and HDPE-specific filaments become more common, the calculus might shift. In two or three years, printing a batch of grid inserts might actually be faster and cheaper than cutting sheet stock. But for now, subtractive manufacturing is the sleeper winner for flat-pack storage solutions.
If you've got a weird prompt of your own — a storage problem, a manufacturing question, a tool you wish existed — send it to us at my weird prompts dot com. We'll figure it out together.
This has been My Weird Prompts. I'm Corn.
I'm Herman Poppleberry. Thanks to our producer Hilbert Flumingtop. We'll catch you next time.
