Recently, I’ve been conducting simulations using the “PMV Comfort Map” and I’ve noticed that the simulations are only conducted at 1-hour intervals, but I would like to conduct them every 15 minutes. I’ve realized that simply adjusting the timestep of the period won’t achieve this, as it will still run every hour. However, I would like to have smoother transitions every 15 minutes. Is there a way to achieve this?
In the example below, I’ll illustrate three different times of the simulation, and you’ll notice that the transition is too abrupt:
The thermal mapping recipes only support hourly timesteps right now. I have hopes of eventually supporting sub-hourly timesteps and the way that I would expose it is through the run period option that you already tried. But adding support for this is a lot of work and so it will be quite a while before you see it in the thermal mapping recipes. Still, it’s good that you posted this since I know that someone like yourself would use it.
In the meantime, the best that I can offer is to use this deconstructed thermal map example here. Because everything in that Grasshopper file is broken down into individual components, you should be able to use it for detailed studies like sub-hourly simulations. You might just have to edit these lines of code inside the Human To Sky Relation component so that you use some decimal hoys to get the sun vectors:
I’ll warn that the deconstructed thermal map runs very inefficiently compared to the thermal mapping recipes in honeybee. So I don’t really recommend it for computing annual thermal comfort metrics. But it gives you a very high level of control and you can even use it to account for some things that the thermal mapping recipes simplify. For example, you can study the impact of shortwave solar falling only on only part of the human subject (thanks to the _height_ and _pt_count_ inputs on the “Human to Sky Relation” component).
The only thing that the thermal mapping recipes have that the deconstructed map lacks is shortwave solar reflections. We can easily account for these in the recipes because we’re using Radiance to model all shortwave reflection but the deconstructed map uses RhinoCommon’s ray tracing methods. So it’s all zero bounces. Generally speaking, the reflected shortwave doesn’t have nearly as large of an impact as the beam and diffuse sun that is being accounted for. So it’s a relatively small drop in accuracy that you’ll have to accept for the increased control the deconstructed map offers.