No Visible Means of Support, part 4 of 4

Turning the corner

It doesn’t occur in the McAlmon apartment, but many of Schindler’s Plaster Skin buildings have two-sided cantilevers, cantilevers that wrap around a corner. Some examples are

Figure 1 Oliver House  1933-1934
front corner above street level garage 1

Figure 2 Buck House 1934
overall front, house on left and apartment on right

Figure 3 Buck House
detail of front at Dining

Figure 4 McAlmon House 1935
front entry of the house (not the apartment) 2

As always, Schindler puts the drama up front. These two sided cantilevers are at, or on the way to, the front door.

Applied to our sample building, a double cantilever looks like Figures 5 & 6.

It is not obvious how to support this two sided cantilever without another beam below it. The answer starts with the single sided cantilever described in the previous articles. Figure 7 shows how this is modified. The front rafter of the single sided cantilever extends, or cantilevers, past the beam that supports it. This extension then supports the second cantilever. In a bit of structural sleight of hand, a cantilever supports a cantilever.

The McAlmon House has the biggest and most dramatic double cantilever (Figure 4). It is about 20′ wide and 8′ deep. Looking at the construction drawings, it turns out that the cantilever is so large that the cantilever beam is a steel “I” section (Figure 8). This is an example of Schindler using steel in a mostly wood building only where absolutely necessary, due to its expense. Note that the steel beam is 7″ tall, slightly shorter than the 2×8 rafters, and fits within the roof depth.

Well, no one’s perfect…

In his efforts to create long, thin building elements and dynamic spaces, Schindler pushes standard framing methods 3. Beams support beams and cantilevers support cantilevers, all with few supports. This is very dramatic. However, I need to point out some weaknesses of his approach to framing.

Standard wood construction has a lot of redundancy-if one piece is a little weak or not cut quite right, the parts around it will compensate. Schindler replaces standard wood walls and few windows with large glass walls. This has the result of concentrating more weight on fewer pieces-it is less forgiving.

The long thin cantilevers are very graceful but Schindler, by current structural standards, undersized much of his framing. It is too thin for the weight it supports and the distances it cantilevers.

Corner and horizontal windows are a common feature of the Plaster Skin buildings. The McAlmon Apartment has a corner window at the front (see figure 1 of article 1 in this series) and wide clerestory windows on two sides of the Living room. This opens up the room, but it also cuts the corner post away from the adjacent walls and the walls away from the roof. This significantly reduces the building’s ability to resist the sideways shaking from earthquakes.

Some final thoughts

In my opinion, this kind of framing and architecture can be done with our current methods. However, Schindler’s approach does need some reinterpretation to meet current structural (and waterproofing) requirements. These changes might include heavier structure, modified proportions to accommodate that thicker structure, strategically placing windows and walls for earthquake resistance, and the use of new technologies that weren’t available to Schindler – such as framing hardware and engineered lumber.


A reinterpreted McAlmon apartment might look like Figures 9-11.

1 Photos 1-3 copyright Steve Wallet, 2010

2 Photo 4 kindly provided by Larry Schaffer
3 Schindler also pushed the limits of other construction technology, particularly the combination of roofs, windows and waterproofing.

2 thoughts on “No Visible Means of Support, part 4 of 4”

  1. I like Buck House, but it's hard for me to imagine that it was from 1934; seems more 60-ish.That must have looked so avant-garde back in the 30's; even today, I think some people would find it "too modern and weird." I want to live in it.

    James in Minneapolis

  2. great analysis and use of sketchup. i agree with your conclusions about the shortcomings especially w/ regard to shear.

    i have mixed feelings about rms' proclivity to conceal everything in an abstract plaster/drywall skin: on the one hand it keeps you focused on the space & forms he's creating, however that missing tectonic dimension is also a missed opportunity.

Leave a Reply

Your email address will not be published. Required fields are marked *