Networks and Engineering Standing Committee Forum

General Topics => Open a Discussion => Topic started by: Toshimichi Otsubo on October 11, 2017, 12:55:37 PM

Title: MPE
Post by: Toshimichi Otsubo on October 11, 2017, 12:55:37 PM
Dear laser ranging colleagues,

On Thursday's session in Riga, we heard a few talks and comments on eye safety issues.  I knew almost nothing about MPE (Maximum permissible exposure) and want to learn a bit.

I was wondering how difficult/realistic it will be to make the energy density low enough (as intended in SLR2000).  But I have a question: in what time window(s) should we worry about our eyes?

I found it hard to understand IEC 60825 even in the Japanese translation, but some of the charts in this wikipedia page is partly understandable (but not sure if trustable).

Let me try to understand by assuming two lasers:
(H) Herstmonceux's:  532 nm, 0.5 mJ, 2 kHz, 10 ps FWHM, beam diameter 3 cm
(S) Stuttgart's:  1064 nm, 0.3 mJ, 10 kHz, 3 ns FWHM, beam diameter 7 cm

(1) Energy per pulse
I guess this chart
  caption: "MPE as energy density versus wavelength for various exposure times (pulse durations)"
is the one to look at.  We can read the MPEs at:
  MPE for (H) = 1e-8 J/cm^2 (where real (H) = 7e-5 J/cm^2)
  MPE for (S) = 2e-6 J/cm^2 (where real (S) = 8e-6 J/cm^2)
so Herstmonceux's pulse is far from eye safety but Stuttgart's is close to eye safety considering the fibre loss.

(2) Energy per longer time span (but how long?)
John Degnan commented that the IR could be dangerous because we cannot see it and the eyes can be exposed for a longer time. 
I am not sure but is this chart
  caption: "Maximum permissible exposure (MPE) at the cornea for a collimated laser beam according to IEC 60825, as energy density versus exposure time for various wavelengths"
useful for this?  Then, we can read at 1 sec (correct?):
  MPE for (H) = 2e-3 J/cm^2 (where real (H) = 0.14e J/cm^2)
  MPE for (S) = 1e-2 J/cm^2 (where real (S) = 8e-2 J/cm^2)
again a similar story as above.

Is my understanding correct?  Can Stuttgart SLR be declared as an eye-safe system by reducing the energy a bit and/or by making the beam aperture large?

Or, if you know the best MPE expert in this community, I would like to contact him/her.

Title: Re: MPE
Post by: jsteinborn on October 11, 2017, 01:23:55 PM
Hi Toshi,

I think what John was referring to:

In general, the human eye will blink within 0.25 second at visible wavelength.
For IR there is no corneal reflex and the exposure time can be much longer.

Title: Re: MPE
Post by: Toshimichi Otsubo on October 11, 2017, 02:06:45 PM
Thanks Jens, I understand that, but I have not understood what graph/equation/number we should refer.  Toshi
Title: Re: MPE
Post by: jsteinborn on October 11, 2017, 02:20:57 PM
This document looks most complete:
Title: Re: MPE
Post by: Johann on October 14, 2017, 01:06:39 AM
<!--                                                                                                           -->Hi Toshi,

to my understanding you are right in all of your points. However, i think there is a bit more than MPE.
Lasers are usually categorized in laser classes. The MPE is similar to laser class 1M, which is safe to the naked eye.
However, if you want an eyesafe laser you have to fullfill the requirements of class 1 (532: max .3,8e-8J in 50 mm Aperture; 1064: max. 3,8e-7J in 50 mm Aperture).
I think the difference between class 1 and 1M is that a class 1 laser is even safe, when using optics with an aperture of up to 50 mm.
ANSI Z 136.1 and i think also IEC 60825 define the requirement for the specific laser classes (enclosures, warning devices, ...).
Finally you also have to take into account the repetition rate of the laser system, which reduces the MPE by a factor of N^-0,25
if the repetition rate is above 600 Hz (IEC 60825-1:2014 p.28).

Concerning the numbers in (A):
Of course it is difficult (impossible) to get eyesafe, or class 1, with a bistatic (small transmitt telescope) SLR-system using picosecond pulse-width.
What our colleagues in Stuttgart do is to use nanosecond pulses to reach class 1M.

To point (B):
I think John meant the "eye-closure reflex", which "should" occure when bright light is seen.
It is assumed that the human eye is closed after a timespan of 0,25 seconds in that case.
Obviously only lasers emitting in the visible spectrum can be categorized in class 2.
I think this laser class is important for lasers with emission duration of more than 0,25 seconds only, not for our ps-lasers.

In Wettzell we are currently working on increasing the repetition rate of the WLRS to 400 Hz. We want to use the whole telescope aperture for
laser beam transmission and we want to switch to 1064 nm. By doing so we are class 1M up to a single pulse energy of 400┬ÁJ with our 10 ps laser.
Recently, we received first light from ETALON1 with an echo rate of about 5%. The approach seems promissing at the moment, we will see whats happening ...

Hopefully i am right with all this stuff, i am slowly getting confused ... ;-)

Title: Re: MPE
Post by: Toshimichi Otsubo on October 17, 2017, 04:07:05 PM

Johann and Jens

Thank you very much.  Amazed to hear that you could track Etalon solidly with such a weak-energy laser.

I am still learning and a bit confused.

I applied the same procedure as Herstmonceux and Stuttgart to the WLRS-IR based on what you wrote.  It does not seem eye-safe in my quick computation.

(W) WLRS:  1064 nm, 0.4 mJ, 400 Hz, 10 ps FWHM, beam diameter 75 cm

The MPE per pulse in the first wikipedia graph is:
  MPE for (W) = 3e-8 J/cm^2 (where real (W) = 9e-8 J/cm^2)
which suggests WLRS-IR is NOT EYE-SAFE.  Am I right?

The MPE per 0.1 s and per 1.0 s in the second wikipedia graph is:
  MPE for (W) = 2e-3 J/cm^2 per 0.1 sec (where real (W) = 4e-6 J/cm^2)
  MPE for (W) = 1e-2 J/cm^2 per 1.0 sec (where real (W) = 4e-5 J/cm^2)
which looks ok.

The N^(-0.25) rule ("Rule 3" in the document from Jens message) will make the MPE simply lower, if I understand correctly.

The PDF document Jens suggested looks useful for visible (400-700 nm) wavelengths - thanks.  I would be glad to see the same thing for 1064 nm.


Title: Re: MPE
Post by: Johann on October 18, 2017, 02:19:54 AM

here are the numbers for the MPE (or class 1M) that i found for 1064 nm:

ANSI Z136.1-2007:
27*t^0.75 J/cm2 ~ 1.5e-7 J/cm2
and N^-0.25

ANSI Z136.1-2014:
2e-6 J/cm2
no reduction from repetition rate

IEC 60825-1:2014:
2e-2 J/m2
N^-0.25 for N>600 Hz

for WLRS this leads to:

~0,6 mJ

2014 (ANSI & 60825):
~8 mJ !!!

it seems that the situation for 1064 nm has relaxed in the latest documents. For all of this numbers the WLRS should be "eyesafe" in the mentioned configuration.

In Wikipedia i found from figure 1: 2e-7 J/cm2
and in figure 3: ~9e-8 J/cm2

However, there is another statement that i found in IEC 60825-1:2014, concerning a Protective Housing:
"Each laser product shall have a protective housing which, when in place, prevents human
access to laser radiation (including errant laser radiation) in excess of the AEL for Class 1,
except when human access is necessary for the performance of the function(s) of the product."

So in principle MPE (class 1M) is not eyesafe. The AEL for class 1 from IEC 60825-1:2014 is:

7.7e-7 J in a 50 mm aperture => ~ 1.2e-7 J/cm2
or ~0.17 mJ for WLRS

There is no clear definition of what needs to be done, when going with lasers through the atmosphere. So it is depending on the corresponding authority, if class 1M is sufficient...

Best regards,
Title: Re: MPE
Post by: Toshimichi Otsubo on October 18, 2017, 03:26:06 AM
Thanks Johann.  Toshi
Title: Re: MPE
Post by: jsteinborn on October 18, 2017, 03:56:29 PM
It looks like we need some kind of math support for formulas :-)

Title: Re: MPE
Post by: danielhampf on November 28, 2017, 11:02:06 AM
Hello everyone,
since our system is mentioned here I thought I should comment. I have just spoken with the laser safety expert in our institute about this. Since I personally know very little about this topic, I will only summarise his comments.
He has checked the above posts and finds many statements which he thinks might be wrong or at least misleading, and therefore dangerous. His main point was: Calculating MPE is a quite difficult business, and almost impossible to do for a general case. Pulse duration, pulse repetition rate, wavelength, beam diameter and divergence are only some of the factors to take into consideration.
We have a detailed analysis of our current system, that defines what precautions are necessary at what point. Indeed, we are currently eye-safe at a distance of a few hundred meters from the transmitter. If we decrease our divergence in the future, this might change. Our procedures will always reflect the specific analysis.
Toshi, since you helped us so much with the data analysis, I suppose we could try to help you with a safety analysis of a proposed system. However, our laser safety officer is reluctant to do public statements in this forum which people might misunderstand, but rely on and possibly do dangerous things with. But if you could send me your specs via PM, I will try to get an analysis from him.
Best regards,
Title: Re: MPE
Post by: Matt Wilkinson on December 21, 2017, 09:55:57 AM
It looks like we need some kind of math support for formulas :-)


Jens, since it is Xmas...

I've installed a mod called mimetex which takes LaTeX and produces a gif image.  It works like this:

[ img]/cgi-bin/mimetex.cgi?f(x)=\int_{-\infty}^xe^{-t^2}dt [/ img]

which without the tag spaces gives:


Please test it out using the post preview.  See here (