This technical paper was written in 1993 before the word "Dosage" was replaced by the correct terminology of "Energy Density."

Measuring UV Dosage Is Not Enough for Process Control

There has been a growing interest in the radiation curing arena in the ability to quantify the curing process. As this interest becomes a need, various instruments have been introduced to measure the UV energy used to cure inks, coatings, and adhesives.

As recently as three years ago, it was thought that measuring the total UV dosage was enough to establish a cure window. By monitoring intensity of the UV exposure over the total time of exposure and integrating that value, it was thought that the UV process had been quantified as thoroughly as possible. As UV users became more sophisticated, they noticed the importance of not only measuring UV but of measuring UV of a certain wavelength. Of particular interest is the wavelength that causes the photoinitiator to become active or "kick over." Also, specific wavelengths of light seem to effect a better cure than others on certain pigmented inks and coatings. For example, long wavelengths appear to be more effective with heavily pigmented inks. Adhesives require both short and long wavelengths to cure properly.

To further complicate the issue, some users believe that the entire UV spectrum plays some role in proper curing. It is not, however, the total of the entire spectrum but the make-up of that spectrum; that is to say, the distribution of the various wavelengths within the UV spectrum that matters. Figures 1-4 demonstrate how different the output spectra of two lamps may be.

Now many formulators and end users are discovering that just monitoring the UV dosage within a particular spectrum of interest is not enough. It is equally important, or perhaps more important, to determine how that UV is delivered. In other words, at what intensity was the dosage received. To give an extreme example, one would get very different cure characteristics by exposing a workpiece to 500 mJ/cm2 of UVA under a 300-W/in. mercury vapor lamp versus laying the workpiece outside in the sun for a period of time which would produce 500 mJ/cm2 (about 3 min). UVA in this example is primarily 365 nm wavelength light.

In the UV curing system, the energy equation would be:

Dosage Energy = UV Intensity x Time
  = 250 mW/cm2 x 2 s
  = 500 mJ/cm2
By contrast, the daylight exposure equation would look something like the following:
Dosage Energy = UV Intensity x Time
  = 2.5 mW/cm2 x 200 s
  = 500 mJ/cm2
In each case the workpiece was exposed to 500 mJ/ cm2 of UVA energy, but the cure properties of each workpiece are very likely to be different. Ignoring chemistry, substrate, pigment and other factors which also affect curing, the controlled UV curing process will produce a more consistent quality product than the uncontrolled, undefined curing process produced by the sun. The point is that measuring UV dosage alone will not assure you of a good cure.

As UV lamps age, the UV intensity decreases. The typical response is to increase exposure time to keep UV dosage (Joules/cm2) constant. This is done by decreasing line speed. While this adjustment does work for a while to maintain desired cure properties, it only works as long as the UV intensity is high enough to effect these properties. This is an oversimplification of a more complex process. Therefore, it is necessary to measure UV intensity along with dosage to correctly characterize the UV curing system.

EIT developed the UV Power Puck in response to a market need for this complete quantification of the UV curing process. The unit measures four separate UV transmission bands simultaneously and displays total UV energy in Joules/cm2, and peak UV intensity in W/cm2 for each channel.

By way of a simple demonstration, the following data was taken to show the output difference between two different UV lamps. Lamp 1 is a standard 200 W/in., medium pressure, mercury vapor lamp. Lamp 2 is an ozone-free version of the same lamp.

Measurements were taken at four different transmission bands: UVA = 320-390 nm; UVB = 280-320 nm; UVC = 250-260 nm; and UVV = 395-445 nm. Though UVV is technically in the visible region, considerable interest has been expressed recently in "UV" curing in this range.

The measurements were taken at four locations along the width of the UV conveyor system running at a constant 25 ft/min. The conveyor is 25 inches wide. Measurements were taken at 5,10,15 and 20 in. from the left side of the conveyor (See Figure 5). Both UV intensity (mW/cm2) and UV energy (mJ/cm2) were measured. The results are shown in Tables 1 and 2.

Several initial observations can be made. First, the intensities on the ends of the both lamps are consistently lower than those in the middle. This is what you would expect and should not come as a surprise.

Secondly, the left side (5 in.) of the conveyor is consistently lower in intensity than the right side (20 in.) for both lamps. It was noticed that the cooling air was being drawn out of the left side of the conveyor at 1800 cfm. It is believed that lower lamp temperature on the left side is the reason for the lower UV output. Graphically, the UV intensity and energy for each lamp is shown in Figures 1-4.

Examining the intensity graphs reveals that Lamp 2 is more intense in the UVA, UVB and UVV regions than Lamp 1. UVC, however, is dramatically just the opposite. Lamp 1 is very rich in UVC and Lamp 2 is noticeably weak in UVC. Remember that Lamp 2 is the ozone-free version of Lamp 1. The graph of Lamp 2 reveals less intensity in the shorter wavelengths known to produce ozone. One might question whether the shorter wavelengths were just filtered out or chemically shifted into the longer part of the spectrum. By checking with the UV lamp manufacturer, it was determined that the quartz jacket of Lamp 2 was specially treated to filter out the shorter wavelength W. A quick check with an EIT UVIMAP indicates that the higher intensities in Lamp 1 are the result of better focusing (See UVIMAP lamp focus curves in Figure 6).

Graphically, the spectral distributions of each lamp are given in Figures 7 and 8.

Since it is possible to alter a UV lamp's output, spectral distribution should be one of the considerations when choosing a UV source. The reaction of the chemistry can be optimized by matching the UV output with the chemistry. Proper lamp and chemistry combinations may be more easily selected for an optimum process by measuring the UV levels for multiple wavelengths with instrumentation that is readily available. After the curing process is characterized, process control may be maintained by monitoring the same parameters on a regular basis.

The new EIT Power Puck provides a rapid method of determining UV intensity and energy at UVA, UVB, UVC and UVV wavelengths. These values can be used to optimize a curing process, monitor a satisfactory process or trouble shoot one which is unsatisfactory.

Table 1. Peak UV Intensity (mW/cm2)
  Distance UVA UVB UVC UVV
Lamp 1
  5 in. 122 120 76 72
  10 in. 136 136 88 78
  15 in. 138 140 90 82
  20 in. 126 126 80 74
Lamp 2
  5 in. 150 138 30 100
  10 in. 160 146 30 106
  15 in. 166 152 32 112
  20 in. 158 144 32 106
Table 2. UV Energy (mJ/cm2)
  Distance UVA UVB UVC UVV
Lamp 1
  5 in. 94.40 102.40 72.45 54.50
  10 in. 106.36 114.00 80.85 62.30
  15 in. 105.65 113.65 81.95 62.10
  20 in. 91.70 99.95 71.45 54.00
Lamp 2
  5 in. 103.65 103.55 18.25 67.50
  10 in. 115.55 114.20 20.10 77.45
  15 in. 115.45 113.05 19.50 77.60
  20 in. 103.90 101.00 17.30 66.95

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

EIT, Inc., 108 Carpenter Drive, Sterling, VA 20164
Phone: 703-478-0700 · Fax:703-478-0815

Questions, problems, comments? Contact
All contents © 1998 EIT Inc. except where noted. All rights reserved.

Last updated