OK, I can't help myself, I have to begin with a final more disclaimers! I'm going to go through my approach with this staircase, and along the way try to pull out the 'rules' or 'principles' driving my approach. In the interest of clarity, I'll just write 'do this' and 'don't do that' and use words like 'always'. However, it should obvious there's no one 'right way', and one could imagine alternative design approaches (e.g. iCopy based 'middle-out' methods) that have a completely separate principles and rules. Further, I am certain that clever people out there have better workflows to do top-down modelling than I do. Having said that, the top-down methods described here are a pretty standard way to achieve model robustness in a reasonably efficient manner. Differences in implementation and technique are perhaps not as important in having a consistent and strategic overall approach.
The image above shows the very first step in designing in this staircase. The lines you see drawn constitute the stair nosing line, running from the bottom finish floor level to the top finish floor level, with a short landing in between. In this case, we know that the total rise (4265mm) is purely nominal, and will change once we get final confirmation from site, so it's vital that everything we do from here can adapt robustly. Note the (500mm) dotted line at the top is the top step, which we will define (rather arbitrarily) as extending to the structural face of the top floor. We always find that being able to adjust this top step independently is useful. We have some user parameters also at this stage, tread going is set to be 280, and the total number of treads, and the number of treads in each flight. This gives us the total going of each flight, and therefore the slope, and the rise of each tread. It's a simple sketch, but it defines the most critical properties of our staircase. Everything else is built around these lines.
Here we can see the creation of some rectangles to mark out the position and extent of the treads. A common tread thickness is used (it can be nominal, or based around our knowledge of the detail of the tread construction, as in this case). Work axes from the first sketch are also introduced.
At this point, we have extruded the bodies of each of the tread markers 1500mm, and are in the process of extruding the top structural floor cavity. We need to nominate the floor coverings and structural floor thicknesses in order to do this. We also need to define the dimensions of the top floor cavity, the dimensions of which we also expect to change upon final site measure.
This is where things start to get quite interesting. We'd like to define the sweep of the structural steel stringer members and we have to consider what defines its location. We have to think about what drives what, i.e. the logical order of elements. The treads are positioned on packers, which are screwed to folded metal supports, which are in turn welded to the top of the stringer. Treads obviously have to remain clear of the stringer. So, we have chosen to draw the basic dimensions of the folded tread support to make this clear. We could well have not done so, and just nominated a clearance between the treads and the stringer, and then fit the tread support subsequently. Mostly, these details were added to help us think about what happened at the landing. We have sketched in a few properties of the tread and riser and this stage too.
To sum up, at this stage we're doing only 10% layout modelling and 90% thinking through our design. Towards the end of the process, we'll have shifted towards doing 90% (detail) modelling and 10% thinking about our design. Right now, we're taking care to introduce elements in their logical order, and being pretty parsimonious with information. Avoid un-necessary details in your 'master parts' like the plague! If we can leave out as much detail as possible until late in the design process, we'll have more flexibility to change our minds, update errors will be easier to fix, and the model will be largely self-documenting. To be honest, we did violate this approach slightly with the treads and riser details, which helped firm up the design in our mind. Strictly speaking, I'd advise working out these details separately from your layout parts. We take the attitude of a classical painter, blocking in the main areas over the entire canvass, and leaving details until the end.
Published on: 03-Aug-2010. Topic/s: 3D design technology