Suppose you want to fully or partially enclose the beam from a Class 4 laser. The combinations of laser/situation/enclosure in question could be anything from (a) a 0.5 W Argon Ion laser, temporary screening during a medical treatment and a lightweight safety curtain, to (b) a 10 kW CO2 laser, operation of a laser welding machine and a steel enclosure. How do you decide what is acceptable to use and what is not? You may also be wondering if there are particular safety standards that the screen or guard has to satisfy and what is available 'off the shelf'.

Up to now the situation has been quite unsatisfactory: no safety standard, no formal procedure for assessing suitability and misleading advertising for high price laser viewing windows and safety curtains. However, a new international standard IEC 60825-4 'Laser Guards', issued November 1997, addresses the design and testing of fixed and temporary guards for laser processing machines but the guiding principles and criteria set out in the standard can be applied to all situations.

There are at least three aspects to the laser guarding problem (and before going any further, please note that the terms 'guard' and 'screen' are both in common use, the former in the machine context only, but they mean the same thing):

• the opacity of the material (i.e. its optical density at the laser wavelength);
• its performance of the material at its maximum level of exposure to laser radiation (e.g. bleaching effects, melting, flammability);
• its general fitness for purpose (i.e. its integrity under prolonged exposure to UV, humidity, temperature variations, wear and tear etc.).

The first and last of these points are straightforward enough to deal with, but what of the second? For example, even a 1 W beam can melt its way through an opaque plastic sheet. It is in dealing with this issue that the benefits of risk assessment over a prescriptive approach to laser safety are most clearly demonstrated.

IEC 60825-4 sets out a risk assessment structure for tackling the problem. At the heart of the standard are two terms: Foreseeable Exposure Limit (FEL) and Protective Exposure Limit (PEL). The FEL, assessed by the user of the guard, is defined as 'the maximum laser exposure within the maintenance inspection interval, assessed under normal and reasonably foreseeable fault conditions'. The PEL, defined as 'the maximum laser exposure for which the guard is specified to prevent accessible laser radiation from exceeding the class 1 AEL', is a term used by providers of laser guards.

In addition to setting out the risk assessment approach, IEC 60825-4 established test conditions and labelling requirements for proprietary laser guards i.e. guards sold for the purpose of providing protection from laser radiation. This includes laser viewing windows which up to now have been misleadingly included in the EN 207 'Filters and Equipment used for personal eye protection against laser radiation', a standard that, as the word 'personal' in the title confirms, applies only to eye protection that is worn or carried. With the imminent release of EN 60825-4, laser machine integrators and users alike should feel more confident about avoiding the use of high cost EN 207 specified materials for viewing windows.

A prescriptive approach to the problem of guarding would involve making somewhat arbitrary assumptions about the maximum level and duration of laser exposure, as the caption above illustrates. Fault conditions are either ignored (the 'false sense of security' approach) or the guard is over-specified to sustain prolonged exposure to the maximum available power of the laser (i.e. the so-called 'fortress enclosure' approach). A third option (the 'protective eyewear' approach ) is to adopt some arbitrary level of exposure and period of exposure, for example that the guard is exposed to the full beam for 100 s, during which time the operator will see that something is wrong and activate the emergency stop. This approach can lead to the worst of both worlds, yet it is the approach of the standard EN 12254 'Screens for Laser Working Places'.

The motivation behind EN 12254 is primarily to produce a simplified standard for manufacturers of laser safety curtains. The problem of testing for high power laser exposure is a serious one. When it comes to laser damage the variables include:

• the duration of exposure (will the early signs of damage be noticed?);
• the repetition rate of the laser, not just the number of pulses to which the guard is exposed;
• the diameter of the incident laser beam (the damage threshold of many materials for CW laser exposure is best measured in Wm-1 rather than Wm-2);
• the distribution of power/energy within the beam ('hot spots' are a particular problem);
• the condition of the guard surface (will it always remain clean?);
• the inclination of the guard and laser beam (horizontal or vertical, normal incidence or close to Brewster's angle).

For the first two of these reasons alone, it is not possible to specify a set of general tests on the basis of which a guard manufacturer can state 'suitable for such and such a power of such and such a laser'. Moreover, there is no way such tests can be reliably scaled down to allow, for example, the use of a lower power laser for test purposes.

Many laser users will not have looked closely at the test conditions for laser safety eyewear as set out in EN 207 (section 4.3 'Stability to Laser Radiation'), but such an examination would reveal a number of gross simplifications, in particular in the treatment of pulsed lasers (where the repetition rate of the laser is ignored) and beam diameter (where the only requirement is that the spot diameter must exceed 0.1 mm). The counter claim, that there have been no documented cases to date where EN207- specified eyewear has failed, is due to the purchaser assuming worst case conditions when specifying the required level of protection for the eyewear, and does not justify the testing which, we would claim, does not generally provide the 10 seconds of protection at maximum exposure that the standard implies.

The issue of IEC 60825-4 is, in effect, a statement by the main international committee of laser safety experts that there is no simple way of dealing with laser safety screens. Would-be purchasers of such screens are strongly advised to follow IEC 60825-4 to determine the Protective Exposure Level they require for the screens and then include the PEL in the test specification for the screen. The safety screen manufacturer is unlikely to have access to a laser the same as the purchaser's, so tests may have to be conducted on site, but this is unavoidable.

Pro Laser provides expert assistance in risk assessment and the selection and procurement of protective filters and other guarding components. We also provide a complete service for the design of laser guarding, including active guards for viewing windows for very high power lasers.

Mike Barrett, principal consultant at Pro Laser, is convenor of the IEC working group that drafted the standard IEC 60825-4, a role that Mike Green held for many years.