design of pavement lights - a grillage or 2-way slab

huangyhg

design of pavement lights - a grillage or 2-way slab
i am having to design pavement-lights for some railway stations and i am having decision problems on a method of design.
basically, the light consists of 200x200mm glass lenses supported of a reinforced concrete grillage with beams 62x125mm at 254mm centres. these beams span in two directions with single reinforcement forming a grid of 3m by 6m approx.
i have no problem with analysing the pavement-light except when i consider the behaviour of the grillage.  the simplest case would be to consider the grillage as one-way spanning slab with primary steel spanning the shortest direction.  the steel in the long direction would be secondary.  but when i consult the german standard for pavement lights (the uk do not have a code), it is advised to design the light as a two-way slab.
my problem is that i cannot see the "wood because of the trees".  i do not think the one-way spanning beam design is the correct method to use (the design will be conservative and safe though) but i am not happy about the two-way slab method either.
has any one any suggestions, please.  i know some lights were tested at imperial college but i have not been able to see any of the results.
regards
hemis
if there's no madatory code saying otherwise and you have truly support at 4 sides, i would design it as two way (maybe i would check the then placed reinforcement was enough to sustain any loads service level, as a safeguard).
in any case i would model also the supporting girders or beams, just to acknowledge the best the actual behaviour. modeling partial loads for delicate structures like this may also be pertinent.
the lesser diameters will show better behaviour inserted in thin   
in the uk, these are usually supplied as precast.
i would just do the one way design and put this reinforcement in in both directions.
the problem with designing it as a two way slab is that you would require torsion resistance in the corners (which you dont have). also if you designed it as a grillage you would also have issues with torsion of the beams.
this is a case where being conservative is possible the cheapest way to go. your time is worth more than it could potentially save.

thanks to both for the replies.
my project involves replacing an existing light that has failed due to water ingress through cracks. frost has blown the underside of the beams exposing reinforcement. the whole is a complete mess.  we are replacing the light using stainless-steel bars and a latex bound waterproof concrete. fibre reinforcement is being added to the 10-mm concrete mix.
i have designed the grid as a series of one-way beams and this is the design i have put forward for checking, but as you rightly point out, as a grillage, there will be torsion at the nodes - maximum in the corners.  this has been shown to me in my computer analysis.  my concern is that the torsion may cause problems.  if i could, i would remove the two side bearings, but i cannot.  i am adding fibre-reinforcement into the mix (plastic or steel) to help tie the mix together but i need to research fibre mixes to see if this would allow the thin-beams to be flexible under torsion.
if any one has any experience they could share with a two-way spanning grillage or fibre additives, i am all ears.
regards
the classical approach that worked well even for walking on was using smaller circular glass elementes embedded in a proportionally quite thick grill of ferrocement. these were cast solidary with the floor and as well as removable precast units on a 4-angle frame.
i agree with csd72 in that if available some industrial product be preferably used according to their specs, they will have debugged all the issues.
respect to differential movement at supports i think you have the answer in structural silicon. it is able to be used even to glue glasswork hanging downwards and also as sole structural fixation for window glass at high exposure to wind, suction of course included, something the codes, wisely, wouldn't allow without some complementary  mechanical fixation. it will acommodate movement whilst providing watertightness and support.
a high end architect used luxcrete on a 10 million pound renovation so they must be good.
if i get this design right, my client will be luxcrete.  the window that is being replaced was built by them. luxcrete in-situ side of the business went out of business.  my client "new age glass" is grabbing the cast in-situ work.
i have decided to replace the short-side bearings with a compressible material, so making the grillage span one-way. i attach a photo of the drawing if you're interested.
thanks for all the comments and advice.
regards
you didn't attach a photo, you attached a link to some site called facebook.
hokie66
i put a photo of the drawing onto face book and the link should have taken you there.  did not work though. if you have a face book account, look up my photo album named drawings. i am putting drawings in there until i get my server to store them.
regards