Prestressing with carbon rovings

Project data

Report in the yearbook 2015

Prestressing with carbon rovings

© Harald Michler

In several projects, the suitability of carbon textiles as reinforcement could be shown. With the carbon reinforcement, extremely thin components and therefore light constructions are feasible, but due to the nature of thin elements a deficit of stiffness needs to be taken into account; extreme deflections can occur. In experimental tests, this seems to be impressive; in practice, such components are useless.

The main reason for such large deformations is the low moment of inertia of thin structural members. Another cause is the cracking behaviour that is similar to that of steel reinforced concrete. Typically, the reinforcement is only activated if a crack has developed. By prestressing of the rovings, the carbon reinforcement will work at a more economic strain level. At the same time, the concrete component remains un-cracked even longer. This also will result in lower deflections. Commonly, the crack load should be > 50 % of the ultimate load. Thus, the component will remain uncracked and therefore stiff during service. Hence, the preloading of the carbon reinforcement has an extremely beneficial effect to a concrete member.

On the other hand, carbon reinforcement is extremely sensitive to lateral pressure, and therefore not suitable to clamp. In the experiment, round bolts (Ø > 50 mm) are used to introduce the load. By turning single rovings around the bolts, they can be anchored by friction. This principle works both with single rovings as well as with carbon textiles, and it was used in a stressing bed for 30 mm thick concrete plates. When using modified textiles prestressed with 30 % of the ultimate load, no significant loss of prestressing force is found in a period of 60 days.

The bending tests have shown very good results. There is an extremely narrow and finely distributed crack pattern. The deflections are very low in comparison to those of a non-prestressed specimen. The largest loss of prestressing force occurred during casting of concrete. The water in the fresh concrete makes the roving weak, leading to relaxation of the prestressing. That isn’t a surprise if one considers the roving as a multi-filament bundle with about 50,000 individual filaments. Improvements can be made by a second coating of the rovings.

Overall, this method seems to have a great potential. Also the use of bigger rovings, with cross-sections of 100 or even 200 K (200,000 filaments), is feasible.