# Significant Result Discrepancy with Parametric THERM Model Tutorial

Hi!

I’m wondering if something is wrong with ladybug/honeybee/therm. I’m currently playing with the sample Parametric_THERM_Model linked from the main website here.

I’m getting results that are different by an order of magnitude for the THERM part of the grasshopper script, and I haven’t changed anything at all in the script. Is this an issue that others who have perhaps used this tutorial faced? If so, how was it addressed. I’ve attached a photo of my output as well. For reference, I’m using Rhino 6 and Therm 7.6 is installed on my computer.

Any help is greatly appreciated,
Ashley

@ashleyh

I think your Rhino document units are in meters, whereas this study needs it to be in inches. Thus your input wall depths are way too high, which increases resistance, and could explain why there’s no actual thermal bridging from the metal studs in your example image

Assuming your document is in meters, and 1 inch equals 0.0254 m, all your wall material depths should be scaled by a factor of about 40 (1m / 0.0254m). If you took out the inch to meter conversion in the “R-Value Without Studs” calculation, I think you might see the R-value increase by about that much (~2000 R) since the materials are laid out in a series and thus are essentially linear systems (with exception of air films?).

The incorrect “R-value With Studs” differs by about a factor of 8 relative to the Hydra example (16.1 versus 141.1), which is a lot less than the factor of 40 derived from the unit conversion. This seems reasonable to me since the addition of the (non-series) studs could account for that factor of 5 reduction in your geometricaly overscaled R-value.

S

2 Likes

Yes converting from feet to inches did it! I assumed feet because the error the components gave upon loading the document mentioned a tolerance issue specifically with feet and I typically work in metric. Thanks for the help Saeran!

Ashley

Great, and thanks for letting me know it was feet not meters, it’s useful to know the result was off by a factor of 12 not 40.