Personalised Drinkware as a Business Line: What a Cylindrical UV Printer Changes

A cylindrical UV printer for personalised drinkware production is usually sold on the wrong number. The brochure leads with print speed. The thing that decides whether the business line works is handling time, and handling time is almost never in the brochure.

That distinction is not pedantry. On a rotary UV printer the physical wrap of a 90 mm tumbler takes a fixed, fairly short amount of time. Loading it into a jig, aligning the seam, unloading it, and swapping the jig when the next order is a 750 ml bottle instead of a 20 oz tumbler — that is where your hours go, and that is what caps how many made-to-order items a day you can ship. Below is the arithmetic for both the throughput ceiling and the money, written so you can drop in your own figures.

Why single-piece orders are possible at all

The mechanism is worth stating plainly, because it is what separates this from screen printing or pad printing a drinkware line.

There is no screen to make, no film, no plate. Artwork goes from a file to the object. That means changeover between one customer's name and the next customer's name is a file change, not a setup — a run of one costs the same per piece as a run of a thousand. And because UV ink cures instantly, the printed object is handling-ready as it comes off the machine: no drying rack, no work-in-progress buffer, no second touch to pack it.

Those two properties together are what make a genuine single-piece flow possible. You can print an item after the order lands and pack it in the same motion. Every conventional decoration route forces you either to hold decorated inventory or to batch orders until a setup is worth doing.

That is the opportunity. Now the constraints.

The throughput model

Per-item time is three terms, not one:

t = T_handle + T_print + (T_jig ÷ B)

• T_handle = load plus align plus unload, seconds per item
• T_print = machine time for the wrap, seconds per item, including every pass
• T_jig = jig changeover time between SKUs, seconds
• B = batch size per SKU before the jig changes

Then:

Items per hour = (3600 × U) ÷ t

where U is realistic utilisation — the share of the hour actually spent producing, after file handling, order picking, packing, breaks and the occasional reprint.

Illustrative placeholders, purely to show the shape: T_handle 25 s, T_print 95 s, T_jig 360 s, B 40 items, U 0.75.

t = 25 + 95 + (360 ÷ 40) = 129 s → items/hour = (3600 × 0.75) ÷ 129 ≈ 21 items/hour

Now the sensitivities that matter, and they are not the ones people expect:

Read the varnish and primer rows twice. A quoted cycle time is meaningless unless you know how many passes it assumes. If the quotation says 95 seconds and your product needs primer plus colour plus varnish for the durability you are promising customers, your real cycle is roughly double or triple that, and your capacity plan is wrong by the same factor.

The money: two separate calculations

Calculation one — replacing an outside decorator

If you currently buy decorated drinkware from a contract decorator, the crossover is a straightforward annual-volume question.

Q*_annual = Annual fixed cost ÷ (D − v)

• D = decorator's price per decorated piece, at the MOQ they enforce
• v = your in-house variable cost per piece = ink + labour, where labour = (t ÷ 3600) × R_loaded and R_loaded is the fully loaded operator plus overhead hourly rate
• Annual fixed cost = machine amortisation + service + spares + space + software

Illustrative placeholders again. Ink $0.55/item. t 129 s, R_loaded $26/h → labour $0.93/item. So v = $1.48/item. Annual fixed: landed machine cost $18,000 over five years = $3,600, plus $900 service, spares and space = $4,500.

Two things fall out. The crossover volumes are low — on these placeholders, under 2,000 pieces a year in most cells, which is a few dozen a week. And the throughput ceiling from the model above (21 items/hour × 6 productive hours × 250 days ≈ 31,000 items/year) sits an order of magnitude above every cell in the table. On a cost-substitution basis the constraint is demand, not capacity, and not capital.

Add one more term if it applies: the decorator's per-design setup charge. If they bill S_out per artwork and you run A distinct designs a year, subtract A × S_out from your annual fixed cost before dividing. For a business with many designs and small quantities per design, this term alone can dominate.

Calculation two — orders the decorator will not take

The table above only describes work a decorator would accept. For a one-off personalised item — a name, a date, a photo — there is no outsourced comparison, because no contract decorator sets up for one piece. So the relevant number is not a crossover at all. It is a floor price:

Minimum viable price = v + fulfilment + (t ÷ 3600) × F_hourly

where fulfilment is blank cost plus packaging plus payment processing plus any shipping you absorb, and F_hourly is the hourly rate you need the machine-hour to earn to cover fixed costs and target margin.

Work it the other way round: take the price you can actually charge for a personalised tumbler, subtract blank, ink, packaging and fees, and you have contribution per item. Divide your required hourly earnings by that contribution, and you get the items per hour you must sell. Compare it to the items per hour the cycle-time model says you can make. If the required rate is far below the achievable rate — which is the usual result — then the business risk sits in marketing and order flow, not in the machine. Plan accordingly, and be honest with yourself about which problem you are actually solving.

Three details that catch people out

• Tumblers are cones, not cylinders. Most popular drinkware tapers from base to lip. A rotary jig that spins a tapered vessel at a constant angular velocity presents a different surface speed to the head at each end, which shows up as stretch or compression toward the lip. Ask specifically how taper is handled — chuck geometry, tilt, software compensation — and ask for the accepted diameter range at both ends, plus the maximum length, rather than a single "diameter" figure.
• A 360° wrap has a seam. Where the wrap closes there is an overlap or a butt joint, and depending on the mechanics it can show as a visible band. Design around it: put the closure under a handle, at a graphic break, or inside a deliberate solid block. Ask to see a full-wrap sample of a solid mid-tone colour — a busy photographic image hides a seam that a flat pastel will not.
• Powder-coat lots vary, and adhesion follows them. Stainless drinkware is usually powder-coated, and coating chemistry and cure differ between blank suppliers and between lots from the same supplier. Ink that bonds beautifully to one batch can fail a fingernail test on the next. Qualify adhesion per blank supplier and per lot, not once at commissioning — and write the qualification into your incoming-goods procedure, not into a folder nobody opens.

A durability protocol to run before you promise anything

Adhesion and dishwasher durability on printed drinkware are exactly the kind of claim that should be tested rather than repeated from a brochure. Run this yourself, on your blanks, before you write a word of product copy:

1. Cross-hatch adhesion to ISO 2409 or ASTM D3359 on a printed panel of the actual blank — not a test coupon. Record the classification.
2. Dishwasher cycling. Run printed items through a domestic dishwasher on a normal hot cycle. Inspect at 10, 25 and 50 cycles for gloss loss, edge lift and colour shift. Decide in advance what counts as a failure.
3. Abrasion. A defined number of double rubs under a defined load, or a tumble test with other items, mimicking a bag or a cup holder.
4. Thermal shock. Fill with boiling water, then with iced water, repeated. Coating and ink expand at different rates.
5. Chemical contact. Hand sanitiser, sunscreen and insect repellent are the three that destroy printed drinkware in the field. Test all three.
6. Repeat the whole protocol on a second blank lot before you commit to a supplier.
7. Write the pass criteria into your machine acceptance test, so the pre-shipment test print is judged against them rather than against how it looks on the day.

What to ask a machine builder next

• Cycle time per item for a full 360° wrap, stated as load / print / unload separately, and stated per pass — with an explicit answer on whether primer and varnish passes are included.
• Jig options: available positions, changeover time between SKUs, and whether quick-change tooling exists for your specific blank sizes.
• Accepted object geometry: minimum and maximum diameter at both ends, maximum length, how taper is compensated, and what happens with a handle in the way.
• Whether a primer or adhesion promoter is required on stainless steel and on powder-coated blanks, and what that does to cycle time.
• Ink cost per item at your artwork coverage, printhead life, and the replacement price list — in writing, before the deposit.
• A pre-shipment test print on your blanks with your artwork, followed by the durability protocol above run on those samples. Same-day sampling on standard models makes this a reasonable thing to insist on.
• Installation, operator training and remote support terms; the after-sales and installation guide covers what to pin down before shipment.
• Payment milestones. Under 30% deposit and 70% before shipment, that balance is your only remaining bargaining power — see paying for industrial equipment from China for how to structure it around an acceptance test.
• Lead time against your season. Personalised drinkware is a Q4 business for most sellers; a machine that lands in November has missed the year. Our Q4 sourcing timeline sets out the backward planning.

If you want a supplier to run that question list against, Longrun Printing Machinery in Shenzhen builds cylindrical UV printers for bottles, cups and tubes with a rotary jig and instant UV cure, sells from a single unit, and can sample standard models the same day — which is the right shape of counterparty for a business whose whole premise is that a run of one has to pay.