> Can you give me technical data and info about LPS luminaries where UF
> is over 0.5? I have not already seen some like this. The problem is
> mostly the size of the lamp. You can shield the light, but when using a
> mirror for the shield, this mirror is in the first place reflecting the
> light again on the lamp. So in that way a mirror not increase the
> efficiency of the luminaire and so the UF stay the same. The shielding
> will decrease the light pollution, but will in these cases not increase
> the UF. With quit complex mirrors this must be possible, but I have not
> already seen that in practice. I think that this can be better put in
> the remarks as a suggestion for further development for this reason.

Friedel, I have no IES or Eulumdat files for good LPS luminaires at hand.

It might be good to put together data for at least all luminaires shown
at Chris's pages,
  Linkname: Fully Shielded LPS Lighting Fixtures
  URL: http://www.nofs.navy.mil/about_NOFS/staff/cbl/LPSnet/LPS-products.html

Maybe, many LPS luminaires are too small to be really efficient. However,
for linear tubes, there is no principal difference to linear fluorescent
tubes, for which good indoor luminaires exist already. And for U-tubes,
there is an analogy with CFL again. It is possible to see something like
two or even four images of the tubes at 60 degrees from zenith, having
over 300 cd/klm there. Light output ratio (LOR) can be then over 0.70.

18W LPS are small enough to have relatively large system of mirrors around
them, even in not very large luminaires. Similarly to CFL luminaires.

Positioning such a luminaire over a centre of a broad street, utilance can
be over 0.7. Then the product, UF would be over 0.500 indeed (I should not
write over 0.5, as it might mean 0.551 at least).

Building a database of ies files, isolux lines, and recommended
positioning of all FS LPS luminaires would be a nice service to mankind...

I've taken but the first example from the above pages, Gardco. It has a
solution with UF over 0.50 (assuming a luminaire on a cable over the
street):

# Source file: lpsa18le.ies
# Luminaire flux =   651 lm,
#                   65.1 % of the bulb flux
# between 75 and 90:  6.7 % of the luminaire flux
#  - this part causes just GLARE in case of road lighting and similar
purposes
# 80deg and above:  max  47.0 cd/1000lm ,  1.8 % of the luminaire flux
# 90deg and above:  max   0.0 cd/1000lm ,  0.0 % of the luminaire flux
# CutOff Type: Full_CutOff

# Illumination of a rectangle of -1.4:1.4:-6.0:6.0 pole heights
#   -- but just within 5.67 pole heights:
#  Per cent of luminaire output falling there:   77.8

What LOR can be achieved for LPS lights, is easy to answer, in principle:
almost the same as for indoor LF. Using antireflection coating on the
bottom glass. Perhaps even using the same grids to confine the light to
the road. I'm not sure about the best LOR for indoor LF with grids. Hope
it can be over 0.66. Then even UF 0.55 is not excluded for LPS over
streets.

cheers,
 jenik