APPROACHING THE PERPENDICULAR

 St Mary, Fratton (Portsmouth) the largest of Blomfield’s churches in my study area. I’m continuing the process of adding indicative tracery where appropriate to better convey the “feel” of the design. It’s my first attempt at Perpendicular style tracery. When I was learning about Gothic architecture as a teenager at Barnsley Grammar School, there were three main periods (I think) Early English, Decorated and Perpendicular.  The names and analysis may have been updated (I don’t really know, not being up to date with the academic world) but I find those three labels a useful starting point for making sense of the world of Gothic and Gothic Revival.

 

Blomfield mostly used Early English on his small to medium sized churches, lots of lancet windows, narrow slots with a pointed top where tracery is not even necessary. Some wider window stray into the Decorated zone with more flowing shapes and floral references carved into the stone. Perpendicular takes a more rectilinear approach with decorative flourishes concentrated in the upper portion of a window. I’m still using short front-to-back sweeps here to guard against distortion and/or error messages should the width and height parameter of the family change.

The trick is to establish the primary grid first, then edit this to suggest the non-rectangular portions of the design. I usually have a second, shorter sweep to maintain the linearity of the grid when add the curved embellishments. This is a gross simplification of course, the actual stonework is a highly three-dimensional, sculpted set of forms.  It would be interesting to do a more detailed study of this, as I started to do a year or so ago, but not within the massing models of these churches which are intended to be viewed within the context of their settings (churchyard, village, town, landscape)

 

I am considering the merits of adding a splayed edge to these windows. One way to do this is to create a void sweep with a triangular profile whose path uses the same geometry constraints as the main void extrusion which defines the window in the first place. It would be possible to use a “pick edge” approach to defining the path, but this is inherently risky because the void sweep is actually destroying that edge by design. Ignore these technicalities if you don’t use Revit. Let’s just say that I’m trying to find low-resolution versions of the forms that make up a church in order to better understand the geometry and overall form.

 

Next come Assembly Views, a quick way of setting up a sheet for a Revit family like this. I already have a view template from previous churches withing the master file, so it’s very fast to set up four elevations and a 3d view (parallel/orthographic)  These views illustrate the logic of taking a broad-brush approach. The finer detail of the tracery is already lost in an impressionistic blob of black lines. Further elaboration would be pointless.

 

 

The 3d view does have a higher resolution and begins to show the limitations of the simplified forms if you zoom in close with a critical eye. But that’s not the intention of this work and it’s always important in the sphere of digital models and BIM to assess the appropriate level of information/detail for the purpose at hand. Are you adding value commensurate with the extra effort involved for everyone concerned, and the load upon the digital infrastructure (computer memory etc)

This work is being done in a Revit file stretched far beyond the size limits recommended for the software. There are more than 400 individual churches within the study area. I’m already struggling to bring in site information via Autodesk Forma.  Let’s keep it simple folks.