Multiple Refractive Surfaces: Too Much Reflection??

[render]

I'm absolutely baffled!

Model:
I'm attempting to render a structure with multiple refractive surfaces. To explain, I'm creating a cylinder like structure in Blender (blend file attached). The cylinder is broken up into partitions along it's height. Each partition is broken up into several torus-like (defined by two circles) rings. Each of the rings--when the project is finished--should have a different index of refraction. For now I've defined them all to be 1.5. Each ring has it's alpha value set to 0. The model is an attempt to discretize a lens with a gradient index of refraction as, to the extent of my knowledge, LuxRender doesn't support materials with gradient indexes of refraction. For testing purposes I've stuck a camera on one side of the cylindrical lens and a plane with a UV image texture on the other. Without the lens the image renders perfectly.

The Problem:
I don't know if this is a model issue, a material issue, a LuxRender issue or a mix of the three. When rendering in LuxRender the edges of the outer rings glow white and the inner rings are not transparent despite my rendering settings (huge eye and light depth). To diagnose the exact problem I reduced the number of partitions in the cylinder I mentioned above. As more of the partitions are added to complete the cylinder the stronger the white rings appear. I've attached a picture of what a one partition rendering looks like. It's possible to see the image on the other side of the lens but white circles are still present. I think the problem has to do with internal reflection of the rings/torus within each partition. I want the light to transmit only through the lens, no reflection. I couldn't find a material setting like this in the LuxRender wiki.

LuxRender Setup:
I'm using Blender 2.49 with the LuxBlend 0.6.1 exporter. Based on the information I've found in the Render Settings section of LuxRender's wiki and information on this board, I've used the following render settings:
Preset: MLT/Bidir Path Tracing
Sampler: metropolis, strength .6
Integrator: bidirectional with eyedepth and lightdepth 2048 (to handle all the possible optical interfaces that the tracer might encounter), strategy is auto, light threshold and eye threshold are zero
Volume Integrator: single with 1.0 as stepsize
Filter: mitchell with sharpness .250

The material settings for each of the rings in each partition are identical. They are:
Material type: glass
Reflection: 0r 0b 0g
Transmission: .8r .8b .8g
IOR: 1.5 (this will change in the final version, but during testing I have all the materials set to 1.5)
All other settings are off

Any help would be appreciated. I've spent forever muddling over this.

[render]

Forgot to attach the blender test file. The Blender file itself was too big to upload (40 mb). I 7zed the file along with the required test image on the UV plane. Both need to be in the same directory.

[heretic]

1. I dont see the reason why you use

Transmission: .8r .8b .8g

on the material? Its totally pointless since it adds a kind of absorption, without doing anything else.
2. Why glass and not glass 2?
3. Who says you cant use an IOR-gradient? For glass just use the gradient texture (uhmmm - ofc, why not?), for glass2 either make your own cauchy values for a smooth transition or sellmayer or tabdata...
4. ....

Sorry I dont get for which problem your approach is a solution since your basic assumption (gradient IOR is not possible in luxrender) is wrong.

[Qantourisc]

What i'm most confused about: what's it with all the rings ? can't this model be LOADS simpler ?

[Eros]

I just looked at the model and wondered the same thing, the model is incredibly complex. Using glass2 would help to make sure the IOR is correct for each section. In normal glass, setting the correct IOR can be tricky Two volumes connected by an interface with the same IOR (say 1.5) should actually have the IOR of that interface with a value of 1. I have found previously that getting this wrong can cause unexpected reflections... when a body that should be transparent suddenly becomes a perfect reflector.

With such a complicated model, I would say using the glass2 interface would help simplify this, the only parts that would need to be changed therefore are any glass/air interfaces, glass glass would be taken care of. I might have a go at this later on assuming i understand the problem (i probably dont however )