Flexo plate distortion: what it is, why it happens, and where prepress shops get it wrong
Flexo plate distortion is the percentage shrinkage applied to artwork along the press direction to compensate for stretch when the plate wraps the cylinder. Here's the formula, the K constants by plate gauge, and the four mistakes that ruin print length.
Flexo plate distortion is the percentage that artwork has to be shrunk along the press direction before the plate is imaged, so that when the plate wraps the cylinder and stretches back out, the printed image lands at the correct repeat length. It’s a flexo-specific quirk — offset and digital don’t have it, because their image carriers don’t wrap and stretch the way a photopolymer plate does.
Get the distortion factor wrong and the printed length is off. A 0.3% miss on a 24” repeat is about 1.8 mm — enough to throw barcode quiet zones, registration marks, and panel-to-panel alignment on a six-panel pouch.
The formula
The distortion factor is:
Distortion % = (K / repeat length) × 100Where:
- K is the plate constant (in the same units as repeat length — usually mm or inches). It encodes the relationship between plate thickness and the stretch the plate undergoes when mounted.
- Repeat length is the press cylinder circumference at the plate surface — most often the gear pitch circumference (gear pitch × number of teeth).
The compensated image length the prepress operator sends to the imager is:
Compensated length = Repeat length × (1 − Distortion / 100)So a 0.625 mm plate (62.5 mils) with a K of 5.0 mm on a 500 mm repeat:
- Distortion = (5.0 / 500) × 100 = 1.00%
- Compensated image length = 500 × (1 − 0.01) = 495 mm
The artwork is built to 500 mm but imaged at 495 mm. Once mounted, the plate stretches the printed image back to 500 mm.
Flexoworks has a free flexo plate distortion calculator that runs this math in millimeters or inches, with K constants pre-loaded for the common plate gauges.
K constants by plate gauge
K is published by the plate manufacturer for each gauge. These are the typical reference values for the common photopolymer gauges — always confirm with your plate vendor’s current technical data sheet, because compound, sleeve, and stickyback choices shift K within a band.
| Plate gauge | Thickness | Typical K (mm) | Typical K (in) |
|---|---|---|---|
| 0.045” / 1.14 mm | 45 mil | 3.3 | 0.130 |
| 0.067” / 1.70 mm | 67 mil | 5.0 | 0.197 |
| 0.107” / 2.72 mm | 107 mil | 7.9 | 0.311 |
| 0.125” / 3.18 mm | 125 mil | 9.4 | 0.370 |
| 0.155” / 3.94 mm | 155 mil | 11.4 | 0.449 |
| 0.250” / 6.35 mm | 250 mil | 18.7 | 0.736 |
DuPont Cyrel, Asahi, MacDermid, and Kodak Flexcel NX publish K constants that line up within ~5% of these for equivalent gauges. The exact K matters — using a 67-mil K on a 107-mil plate produces a 0.6% length error, which is visible on most jobs.
Why plates stretch in the first place
A photopolymer plate is imaged flat. When it’s mounted to a stickyback on a cylinder, the outer surface (where the printing dots live) has to travel a longer arc than the back of the plate. Thicker plates have more material between back and surface, so the stretch is bigger.
That stretch is what the distortion factor compensates for. Imaging the plate at a slightly shorter length means that once it wraps the cylinder and the outer surface stretches, the printed image is the correct length.
The relationship is roughly linear with thickness, which is why K scales with gauge. It also depends on the sleeve diameter (smaller cylinders = more curvature = more stretch), which is why some shops maintain different K tables per press.
The four places shops get distortion wrong
Distortion calculation itself is simple. Where shops lose money is upstream and downstream of the formula.
1. Wrong K for the plate that actually got imaged. Prepress quotes a job for a 67-mil plate. The plate room is out of 67-mil compound and substitutes 107-mil for the same job. The original distortion factor was 1.0%; on the substituted plate it should be 1.58%. If nobody re-files the job, the printed repeat is 0.58% short — about 3 mm on a 500 mm repeat. The press operator catches it at first pull and the plate gets thrown.
2. Repeat length measured from the wrong reference. The gear pitch circumference and the printing cylinder circumference are not the same. Mounting tape, plate thickness, and stickyback all sit between the gear and the print surface. Most flexo prepress is based on gear pitch — but some shops use cylinder OD, some use the published “effective repeat” from the press OEM, and some mix references between jobs. A 0.5 mm discrepancy on a 500 mm repeat is 0.1% distortion, which compounds with K-table error and shows up at the press.
3. Compound or sleeve changes that shift K. A new plate compound from the vendor — different durometer, different relief, different floor — can shift K by 3–8% relative to the legacy compound, even at the same nominal gauge. Same for swapping from a hard sleeve to a soft sleeve, or changing stickyback brand. Shops that don’t re-measure K after a vendor change print short or long for weeks until somebody figures out what changed.
4. Distortion applied twice or not at all. This sounds avoidable, but it happens regularly when artwork moves between the brand’s design agency, the trade shop’s prepress, and the converter’s in-house prepress. If the designer pre-distorts the file (“we always send you files at 99% so you don’t have to”), and the trade shop also applies distortion, the printed image is short by twice the factor. If both parties assume the other one is doing it, the printed image is at full length and doesn’t fit the repeat. The fix is a documented handoff convention per customer, enforced at intake — not a verbal habit.
What distortion does not compensate for
Distortion is a length compensation along the press direction. It does not fix:
- Lateral (across-web) bowing. Bow comes from cylinder runout, plate compound deformation under impression, or non-uniform mounting. It needs press setup and mounting fixes, not a different K.
- Dot gain. Dot gain is a tonal issue — dots grow on press relative to the plate. It’s handled by screening, plate type (HD Flexo, Crystal, AED), and press curves.
- Registration drift. Lateral and circumferential registration between stations is a press-control issue.
- Substrate stretch. Some films and unsupported laminates stretch under web tension. That’s compensated at the press, not the plate.
Conflating any of these with distortion is one of the more common ways junior prepress operators waste plates.
Where the distortion factor lives in the trade shop
The distortion factor sits at the front of the prepress workflow, but it’s a touchpoint between three groups:
- Prepress operators apply it at step-and-repeat, based on the job ticket.
- The plate room has to confirm the plate gauge actually imaged matches what the ticket said. If they substitute gauge, the prepress file has to be re-RIPped with the new K.
- The customer’s press team has to confirm the repeat reference (gear pitch vs. cylinder OD) on file with the trade shop.
In shops running on spreadsheets, each of these touchpoints is a verbal handoff. In a flexo-native MIS, the plate gauge, repeat reference, and K constant are first-class fields on the job ticket — so substitutions trigger a re-distort instead of a silent length error on press.
For more on this:
- Flexo plate distortion calculator — free tool with K constants for every common plate gauge, in millimeters or inches.
- Prepress automation software for flexographic trade shops — where step-and-repeat and distortion sit in the workflow.
- Prepress workflow software vs prepress MIS — the production tool that applies distortion vs. the system of record that tracks the plate-gauge spec.
- Flexo prepress software for trade shops — the MIS that keeps plate gauge, repeat reference, and K constant on the job ticket.