Chris: R = 1.5 certainly feels like a conservative value. I'm confident that an unsheathed timber frame can dissipate a good amount of energy under reversible loads, so strength isn't the real issue. Serviceability under such loads would then become the problem.
The obvious consequence of R = 1.5 is that seismic loads become quite large when elastic or nearly elastic response is assumed. In general, I'd be surprised if a conventional timber frame that relies on knee braces for lateral load resistance could be designed to resist the required seismic loads and satisfy the corresponding drift limits. This suggests that an alternative framing concept (such as story-high bracing as used in German Fachwerk) might be in order. -- Dick

Dick, I have a pavilion type frame that is one story with no walls. The joinery will be traditional timber joinery except there will be cable cross bracing in between the bents in the longitudinal direction and timber cross bracing within the bents. However, this bracing is only in the top quarter of the frame height. From there down there is no bracing. This creates somewhat of a rigid frame effect in that it places the posts in bending at the lower level of the bracing.

Since the timber frame is the only weight in consideration for the seismic loading, no floor, no walls & light roof, the seismic loads should be fairly low.

I don't think this will fall into a moment-resisting frame or a braced frame, especially since the IBC references steel moment frames and braced frames.

I know you are reluctant to recommend an R value for traditional timber frames, but this one is different in that it is partially braced with cross-bracing at the frame. Any suggestions?

Thanks

Rob: My first approach would be to do the seismic load calcs using R=1.0 to determine how that would compare to the required wind load. If seismic loads for R=1.0 are really low, realive to wind, then you should be able to design for wind without concern for energy dissipation. If this project requires a building inspectors approval, then he or she should be willing to accept your PE stamp in place of an R value in the code. If seismic loads control and are unrealistically high, then you need to think the design through in more detail. One approach to carrying high tension through knee braces (and to reduce the negative effects of joinery in the post), is to route a channel in the top of the brace, lay in a steel rod or cable, and then cap the channel. Run the rod throught the post and the receiving girt and anchor with bearing plates. Then use stub tenons to keep the braces in place. That still doesn't give you a structure with an approved R, but it will carry more tension than a few pegs. -- Dick

I spoke with Philip Line of the AF&PA today regarding seismic design with Timber. The current seismic codes originated in California with the help of the Federal Emergency Management Agency (FEMA) and the National Earthquake Hazard Reduction Program (NEHRP). Their main focus was with common types of buildings that experienced problems in earthquakes.

The 1997 UBC has values for Braced Heavy Timber as part of a bearing wall system (R=2.8, Omega=2.2) and as part of a braced frame system (R=5.6, Omega=2.2)

Apparently these or similar values will eventually make it back into ASCE 7 and then the IBC. It may take a couple of code cycles to accomplish it though.

It seems that these numbers are the best bet at this time. In most cases wind will likely govern anyway.