Outdoor glare from facade glazing

grasshopper honeybee
,
1

Hi @Hamzajh19, @MingboPeng , @LittleBuddha, @sarith, @LiamRuvio, @Grasshope
First of all, sorry to bother you, but I’m hoping someone can shed some light on this.
I’m evaluating glare risk from a glazed façade. In my study, drivers act as receptors and the glazing is the glare source. I’ve already identified when and where glare is geometrically possible. The goal now is to select a glazing that mitigates glare risk, so I need to vary the glass reflectance and transmittance. I find this second part rather challenging.

I’ve been using the Glare Hazard plot and the SGHAT manual from Sandia, but my results don’t seem consistent. I’m modeling the driver’s viewpoint as a point with a sight vector.

image
image778×384 36.9 KB

From the Glare Hazard plot, I’m using the retinal irradiance thresholds; above 0.1 W/cm² there’s potential for after-image, which I’m treating as my limit. This corresponds to 12.5 W/m² of corneal irradiance (incident at the eye). However, when I run an annual radiation analysis with Radiance, it’s strange that values below this threshold are rare, does this imply a constant risk of after-image whenever the source is visible?

This is confusing, and I haven’t found a single indicator in Ladybug to quantify this. DGP is intended for discomfort glare in offices and doesn’t seem suitable for outdoor glint/glare, which is more of a safety (disability glare) issue.

I also found the 20,000 cd/m² source luminance limit in EASA guidance. How can this be measured in LBT? A fisheye rendering yields many pixels with different luminances; if I simply take the maximum when the source is visible, it often exceeds this value.

Has anyone addressed this? Which standards or reference criteria are you using for assessment?

Thanks

2

I really like this study from Alstan Jakubiec and Christoph Reinhart: https://journals.sagepub.com/doi/10.3141/2449-13 . The limit they discuss is 30K cd/m sq.

Full text link in case you don’t have direct access.

I also found the 20,000 cd/m² source luminance limit in EASA guidance. How can this be measured in LBT? A fisheye rendering yields many pixels with different luminances; if I simply take the maximum when the source is visible, it often exceeds this value.

Both perceptually and mathematically, the idea of glare is based on certain regions of the field of view being exceedingly bright compared to other regions. So, taking maximum is fine I guess, but it might be a little more reasonable to pick some X% of the total brightest pixels and see if their luminance exceeds a threshold. Perhaps there is someway that evalglare does this automatically, but if not pvalue + some light post processing should do the trick.

3

Thanks for your response. I’m already familiar with that paper, but I think the analysis I need to conduct differs somewhat. In my case, I have to compare several glazing alternatives to determine which ones, if any, can reduce the risk of disability glare caused by reflections. I have already identify when and where this glare is geometrically possible (glazing acting like a mirror), but I’m struggling to assess the intensity of the glare and determine the threshold at which it becomes excessive.