Despite all being less than 100 metres tall, the world's tallest timber buildings all utilise concrete to increase their mass such that they do not vibrate excessively under wind loading. Wind-induced vibrations must be minimised to ensure that the building's occupants remain comfortable and do not regularly experience motion sickness during high winds. Despite the difficulties with wind dynamics for the current generation of timber towers, numerous concept designs have been announced that propose to build much taller with timber. However, at present, there has been little consideration of how the architecture of timber towers can be suitably designed to help combat the problem. This thesis investigates the effects of different structural typologies on the dynamic performance of timber buildings by studying four iconic skyscrapers; the Gherkin, the Shard, the John Hancock Center and 432 Park Avenue and examining how they would perform if built from timber. First, they are assessed at their existing heights and across a range of shorter heights, with their steel or concrete frames but examining the effect of replacing their concrete floors with CLT. Secondly (and again across a range of heights), the buildings are redesigned with a timber frame to test how their dynamics would change if their steel or concrete beams, columns and walls were replaced with glulam sections and CLT panels. The Shard and 432 Park Avenue, which have concrete cores, have also been examined to see how they perform if they kept their concrete cores, but if the remainder of their structures were built from timber.

In total, 144 combinations of building, floor material, height, and frame material are assessed. Retaining their existing steel or concrete frames but replacing their concrete floors with CLT resulted in the buildings' natural frequencies increasing by an average of 30% and the peak accelerations by 47%. These changes are due to the CLT floors being considerably lighter than the original concrete floors. By comparison, the change from a steel or concrete-framed structure to a timber-framed structure (with no change in floor type) made little difference to the peak accelerations, but caused natural frequencies to increase by 11%.

If their existing structures were retained, but CLT panels with a thin layer of concrete screed were used for their floors (instead of deep concrete slabs), then the Gherkin at 182 m, the Shard at 200 m, the John Hancock Center at 196 m and 432 Park Avenue at 137 m would have acceptable vibrations (for residential occupancy) if located in a low wind speed environment like London. Across the four buildings, this change in floor type would save an average of 24 $kgCO2$ per $m2$ of floorspace if sequestered carbon is excluded, and 170 $kgCO2/m2$ if sequestered carbon is included. When sequestered carbon is included in the calculation, the net carbon stored in CLT is enough to offset the embodied carbon of the steel and concrete of the Shard (at 200 m) and 432 Park Avenue (at 137 m). When sizing the columns and diagonals of the Gherkin and the John Hancock Center, the strength criteria was the limiting factor (rather than stiffness). This is because both towers have well-braced tubular designs that are inherently stiff, thanks to the majority of their columns and diagonals being located on their perimeters. With strength as the governing criterion, the size of the structural members could be reduced when lightweight CLT floors were used instead of concrete. For example, the columns of the Gherkin would have required 32% less steel if CLT floors had been used instead of concrete decks. Such savings would not be possible for the Shard or 432 Park Avenue, where the stiffness criterion limits the sizes of the sections.

If the four skyscraper designs were built with a timber frame, the Gherkin would comfortably be the best performing structure thanks to its inherently stiff diagrid shell and its circular cross-section. It could easily satisfy the ISO 10137 human comfort criterion for residential occupancy in most locations at its full height of 182 metres. Taller versions of the structure are also likely to be viable. If built in London, a fully-timber Shard at 134 m (or 200 m with a concrete core and glulam frame), a timber John Hancock Center at 196 m, and a fully timber 432 Park Avenue at 80 m (or a hybrid at 137 m) could also satisfy the same criterion (all with CLT and screed floors). Across a set of the 135 m versions of the four skyscrapers, the change from a steel or concrete frame to a glulam and CLT structure would result in a saving of 130 $kgCO2/m2$ (including sequestered carbon) or a saving of 92 $kgCO2/m2$ for a hybrid (timber beams and columns, but retaining a concrete core).

Overall, when different typologies were compared on a like-for-like basis, braced tubular forms like the Gherkin and the John Hancock Center worked the best in timber, producing lower wind-induced vibrations than 432 Park Avenue and the Shard. Furthermore, their tubular structures required smaller column sizes (which occupy a lower percentage of their floor space), have lower material costs per $m2$ of floor space and would result in less embodied carbon per $m2$ (if sequestered carbon is ignored) than those which rely on an internal core for lateral stability.

The next generation of tall timber buildings looks unlikely to reach some of the super tall heights proposed without significant additional damping, added mass or suitable aerodynamic cross-sections that can minimise wind-induced vibrations. However, this thesis has shown that timber does have the potential to be used in suitably designed tall buildings up to at least 200 m tall, without additional damping or mass, and as the primary structural material in the frame or as an alternative to concrete floors.