It’s January. Outside the window, the snow is falling. On the schedule: a reprint of 400,000 impressions for your biggest customer. You order the sleeves from the warehouse and when it comes time to print, the pressman calls. He can’t get the job to register and you know the last time the job was on in October, it printed flawlessly. This is called thermal contraction.

You walk down to the press and examine the results. Baffled, you lean against the cold metal girder as it stretches from the ceiling to the floor. You wonder, “What is going on here?” And then it dawns on you…

Temperature affects printing. One of the biggest ways is through the expansion and contraction of polymer and elastomer plates and sleeves. Material changes across temperature ranges can produce registration and quality issues with your flexographic printing process, which has tolerances in the thousandths of inches. Knowing how different materials change with temperature is the key to planning a printrun and mitigating mounting, registration and quality issues.


The rate that a solid material changes in length with temperature change can usually be calculated as the fractional change in length per degree of temperature change. This number is called the linear coefficient of thermal expansion and is often expressed in 10-6 m/m K (or degrees Celsius), or in the U.S., degrees Fahrenheit. As long as one stays in the same units for their fraction, it becomes possible to convert this to inches and thousandths of inches easily. It is also important to note that, with solids, the effects of air pressure differences on thermal expansion are negligible.

But the theory becomes complicated in the flexographic printing process, because printing plates and sleeves are not uniform solids. They are compound materials bonded to other compound material. Sleeves are usually composed of elastomer or photopolymer compounds layered around a fiberglass core, bonded to the core with adhesive. Plates are usually photopolymer plates mounted to fiberglass cores using stickyback cushion tape. Because fiberglass formulations generally tend to have a lower coefficient of thermal expansion than most elastomer and polymer formulations, the materials mounted around them cannot expand and contract without strain effects of being attached to the fiber cores.


infrared image of a sleeve on the printer

Further complications arise because different formulations of elastomers, polymers and fiberglass have different thermal expansion characteristics, and it may be impossible to characterize every type of material bonded to every type of core. In addition, not all materials are isotropic (they do not expand at the same rate in every direction).

Often, the formulations and thermal expansion characteristics of the materials are trade secrets held by the different manufacturers.

Because of the difficulty of studying every single type of compound and core, we decided it was important not to study these separate materials in a lab, but to study the effects of thermal expansion on plates and sleeves in real world scenarios: on warehouse floors, in storage rooms, and mounted on engravers and presses.

These four thermographic pictures, taken in three minute intervals, depict sleeves being placed on a cold warehouse floor. Notice how, within three minutes,  the bottoms of the sleeves start to get colder
These four thermographic pictures, taken in three minute intervals, depict sleeves being placed on a cold warehouse floor. Notice how, within three minutes,  the bottoms of the sleeves start to get colder. The heat transfers away from the sleeves via conduction (direct contact between solids) with the cold floor, a much more efficient method of thermal transfer than convection (heat transfer from a liquid or gas to a solid). This illustrates why it’s a bad idea to storesleeves in direct contact with a cold warehouse floor.


For the purposes of this study, we standardized on 48-in. sleeves of different elastomermaterials and on 50-in. photopolymer plates mounted to a fiberglass core. In addition, given the difficulty of measuring diameters of bendable plates and sleeves not on a mandrel, we decided to only measure sleeve lengths at this time.

The outside temperature was around 29 degrees Fahrenheit, so cooling was as simple as leaving materials outside for a couple of hours. Materials were first acclimated for 24 hours off the ground to insure they were the same temperature head to tail. Then the following procedures were performed, and temperature and length were measured at each step:

  • Engrave the material with a measurement
    scale in the material 61-in. long
  • Set material outside for two to four hours
  • Set material inside and allow to acclimate
    for two hours
  • Allow material to sit overnight
  • Heat sleeve

All length measurements had a tolerance of plus or minus 0.002-in. All temperature measurements are accurate to within plus or minus 3 degrees Fahrenheit.


Several trends were readily apparent. First, 90 percent of elastomer is a formulation of EPDM (ethylene propylene diene monomer). We found negligible differences in the variations of EPDM sleeves with respect to thermal expansion and contraction. They performed very similarly to each other.

Second, we noted that polymer versus elastomer plates had very different coefficients of thermal expansion. Polymer tended to be more sensitive to thermal expansion than elastomer. Below room temperature, a 61-in. polymer plate mounted to a core tended to contract (and/or correspondingly expand) around 0.003-in. per degree Fahrenheit, while a 61-in. elastomer tended to contract around 0.001-in. per degree Fahrenheit. Standard 0.25 undercut was used. Polymer plates had a 0.015-in. stickyback, a 0.040-in. base and a 0.067-in. plate.


Third, the thermal expansion characteristics of sleeves and plates did not appear to be linear across all temperature ranges. What we noticed was that from 29 degrees Fahrenheit to room temperature (72 degrees Fahrenheit), thermal expansion of plates and sleeves appeared to be fairly consistent. However, as materials were heated above room temperature, we found they did not expand at the same rate they contracted. Above room temperature, a 61-in. polymer mounted to a core tended to expand 0.0015-in. per degree Fahrenheit, while elastomer tended to expand 0.0005-in. per degree Fahrenheit. The expansion numbers were not nearly as consistent as the contraction numbers.

Fourth, sleeve sizes of 48-in. or more seemed to expand the same per degree as sleeves of 51-in., 53-in. and 61-in. This supports a hypothesis that after a certain length, elastomer or polymer adhered to a fiber core, has finite expansion capabilities and, with elastomers, the rate of contraction is 0.001-in. per degree Fahrenheit, regardless of whether a sleeve is 51-in. or 61-in. (20 percent longer). This is clearly an area that requires more study, as much of this was based on rough estimates.

The previous two observations would support a hypothesis that the strain of the material being bound to the fiber core has a limiting effect on how much the elastomer or polymer can contract or expand, and has a limiting effect above and below room temperature.


Another observation that was readily apparent, was that it took much less time for polymer plates and sleeves to return to ambient temperature than it did for elastomer sleeves. This is due to two reasons that both have to do with Specific Heat Capacity (commonly called “Specific Heat”)—the amount of heat per unit of mass (or volume) required to raise the temperature of a material by one degree. Specific heat is usually measured in joules per kilogram per degree Kelvin, J/kg-K, or kilojoules per kiloliter per degree Kelvin 103 J/m3-K.

The first reason polymer plates contract and expand more quickly is that polyester, the base for most photopolymer sleeves and plates, has a lower specific heat than EPDM elastomer (1 kJ/kg-K versus 2 KJ/kg-K). It takes more energy to increase the temperature of elastomer. Again, this is general, as EPDM and photopolymer formulations for flexography are proprietary.

The second reason is that elastomer sleeves usually have much thicker walls, which means there is more material to return to temperature. As can be seen from the formulation for specific heat, the more material you have, the more energy it requires to raise its temperature—it takes longer to return to room temperature.

In our observations, it took most elastomer sleeves 12 hours to come back to spec, while it took polymer only three hours. Fiberglass cores took two hours to come back to spec. It is important to note that “back to spec”
means the material returned to its room temperature dimensions or file dimensions. We did not base this timeframe off a measure of temperature, because only the surface heat of the elastomer sleeve could be measured. We believe that a “return to spec” meant that the sleeve’s core temperature, not just the surface reading had returned to 72 degrees Fahrenheit. More study is needed and ways to measure the core temperatures of sleeves is something that manufacturers may want to examine.

sleeves cooling down. Great example of thermal contraction
In this time lapse, the sleeve was photographed every minute for 40 minutes, cooling down to room temperature via convection (the air of the room in contact with the solid sleeve). Note that the ends cool faster than the middle, and that even after 40 minutes, the sleeve is not down to room temperature.


There is still much to learn when it comes to how thermal expansion affects polymer plates and sleeves. We feel that this rough study is just a starting point, and that extensive research needs to be done to improve our knowledge of all the effects of thermal expansion and contraction in flexography.

Accurate linear and volumetric thermal expansion and contraction statistics should be provided by plate, sleeve and core manufacturers,
and it is up to end users to make manufacturers aware of the issues. Studying these issues in the lab will help increase the real world knowledge for engravers, pressmen and production managers.

Despite the need for increased communication, manufacturers are constantly making strides. New materials are continually under development. One elastomer we studied had one third the thermal coefficient of expansion as EPDM, putting it very close to the thermal expansion characteristics of the steel bases of rotogravure. We are also working with manufacturers and printers on methods of acclimatizing plates and sleeves more quickly, and are exploring the concept of using embedded devices to report core sleeve temperatures.

Knowledge is power. Knowing how temperature affects the printing
process allows us all to continue to capitalize on the economic advantages of elastomer and polymer over rotogravure, and the run length benefits of elastomer. We look forward to continuing to study thermal expansion issues and working closely with manufacturers to best address them.

No Comments Yet.

Leave a comment