Implementing an innovative renovation concept by using force and deformation measurement technology
The original, historic load transfer for the cupola was no longer effective and instead the load was redistributed to the ceilings and joists of the floors below, resulting in these areas becoming overloaded. The plan that emerged was to restore the historical load transfer of the cupola load. This involved developing a concept that centered on a complex redistribution of the cupola load from the posts to the struts of the truss frame.
1. Introduction and problems
The Schloss Friedenstein is the biggest palace in Thuringia dating from the Thirty Years' War (1618–1648) as the foundation stone was laid in 1643 by Duke Ernst I. It is also the biggest early baroque palace park anywhere in Germany.
Andreas Rudolph, a fortress builder from Magdeburg, was in charge of the construction and the building was completed in 1654.
Fig. 1: Schloss Friedenstein as it looks today
(photo: Ingenieurbüro Hirsch, Erfurt)
The palace was mainly used for residential purposes and official functions, as well as its use as a royal seat, from the baroque period to the age of classicism. At present, it is mainly used by the Gotha Research and State Library and by the Gotha Thuringian State Archive. It is also an important tourist attraction. The south-west corner of the palace courtyard is formed by the west tower, which is 26 meters along the edges and 40 meters high.
The tower houses the Erkhoftheater, which dates from around 1681, and is believed to be one of the oldest surviving baroque theaters fitted with the original stage machinery.
The roof of the west tower is a tent-shaped structure with an octagonal cupola (a dome-shaped construction) mounted on top. The original architectural style used truss frames to divert loads from the cupola to the exterior masonry and created a tower that very spacious and virtually support-free. Truss frames are supporting structures that transfer the forces through the diagonal struts.
Fig. 2: View of the west tower
Fig. 3: Cross-section of the west tower
Over time, damage to the supporting structures, and to the roof’s supporting points on the exterior masonry, had led to a pronounced sag in the ceilings on the second and third stories of the building. The original, historic load transfer for the cupola was no longer effective and instead the load was redistributed to the ceilings and joists of the floors below, resulting in these areas becoming overloaded. Figure 4 shows the planned transfer of load on the left and on the right, the variation in the flow of force caused by the damage.
Path taken by the force as originally planned
Actual path taken by the force
Fig. 4: Comparison of the paths taken by the force (force flow) in the truss frame area
2. Renovation concept
Both the planners and renovators examining the reconstruction task were limited by a number of constraints. They had to:
- find a solution suitable for an historic monument
- avoid the use of additional structural reinforcement on operational floors
- carry out renovation while the building was in use
- retain the historical fabric of the building
- ensure any proposed reinforcement was localized and minimal
- optimize the cost of the renovation
Faced with these constraints, and after weighing up various alternatives, the plan that emerged was to restore the historical load transfer of the cupola load. This involved developing a concept that centered on a complex redistribution of the cupola load from the posts to the struts of the truss frame.
To achieve this it required toughening the wooden truss frames and then fitting eight tensioning elements with integrated force measurement into the diagonal struts. To activate the truss frames, the previous force flow had to be redirected. A diagram of the procedure to be adopted is shown in Fig. 5.
The load from the truss frame posts were temporarily redirected to eight pairs of hydraulic presses, so that the wall plate and the masonry base block below it were without load and could therefore be dismantled. A combined force/displacement measurement was taken at each press location, in order to monitor the load-deformation behavior. This made it possible to prevent additional loads being applied to the area below the wall plate, which was already greatly overloaded. When the hydraulic presses were retracted, the load on the ceiling would be removed and transferred to the exterior masonry via the struts of the truss frame. For this to happen, the load distribution between the individual struts had to be defined first.
The force flow redistribution could only be achieved if, as part of the reconstruction work, the forces and deformations that occur at throughout the truss frame were measured with high precision and displayed on-site.
Fig. 5: Redirection of the force flow
diagram at the top, with the on-site construction situation below
It was also important for the metrological monitoring of force flow and deformation behavior to continue once the renovation was completed. Consequently, a system comprising a combination of electrical online measurements and geodetic monitoring was developed and implemented to determine the deformation states.
3. Areas of application for measurement technology
Measurement technology had an important role to play in the renovation concept for controlling and monitoring the complex processes of load redistribution. This applied to various areas.
3.1 Measuring the jacking forces on the truss frame posts
During temporary load distribution, the force applied by each pair of presses was recorded by means of a force transducer positioned in the force flow. Type C6A sensors from HBM with a maximum capacity of 200 kN were used. These were connected to an HBM Spider8 amplifier with a SR55 expansion module added to increase the number of carrier frequency measurement channels to eight. Force was only measured on one press of each of the press pairs as it could be assumed, from the hydraulic coupling of the two cylinders, that the load was evenly distributed. This type of measuring point is shown in Fig. 6.
By deliberately triggering the individual press pairs, it was considered possible to achieve an even distribution of load to the hydraulic cylinders. There were also specific limit loads that were not to be exceeded because this could overload the ceiling construction at certain points, resulting in the renovation work could be causing more damage than it repaired!
The working method developed was also intended to metrologically define the previously unknown original total loading of the cupola and make these forces available for any additional static calculations on the supporting structure.
The measurement data obtained from measuring the hydraulic force was made available in real time to the structural engineer on site.
Fig. 6: Hydraulic press with force transducer and displacement sensor on the base plate of a truss frame post
3.2 Measuring the amount of lift at truss frame posts
In parallel with measuring the hydraulic forces, the elevation of the truss frame posts relative to the underlying ceiling construction of the third floor was metrologically recorded. This was achieved using potentiometric displacement transducers connected to a separate eight-channel Spider8 amplifier. The measuring points were located next to the hydraulic presses as shown in Fig. 6.
The measurement also made it possible to determine how much of the total loading had been taken on by the hydraulic presses.
This was indicated by the truss frame posts becoming free of load from the underlying joists, which was reflected in the force-deformation characteristic of combined force/displacement measurement. The hydraulic presses were also controlled manually, from the evident displacement.
3.3 Determining strut forces by measuring the strain on the truss frame struts
Another important part of the metrological aspect of the renovation work was measuring the compressive forces in the truss frame struts. This was because the planner had called for a specific load distribution to be realized between the individual struts of the supporting structure. This was made possible by selective stressing during fine adjustment. To prevent overloading, it was also essential to adhere to a specified, safe limit load. The load situation in the truss frame was required to be monitored for a longer period after the renovation.
Fig. 7: Diagram of the combined hanging truss/truss frame with tensioning elements in the diagonal struts
In view of these requirements, a combined measuring and tensioning element (MSE) was designed, and installed in the force flow of each truss frame strut. A diagram of this is shown in Fig. 7. The length of the element could be varied by a thread, allowing the effect on the amount of load on the struts to be controlled. This was done manually.
Fig. 8: Tensioning element with a strain gage full bridge test object at the top, installation in the truss frame below
The same element was also used when measuring force, as the strain was measured by means of a foil strain gage and a relevant force calculation was performed using the linear stress-to-strain ratio. The measuring points were two HBM type 6/120 XY 31 T-rosettes sited on opposite sides of the perimeter. The strain gages were electrically connected to the full bridge. The intention here was to compensate for changes in the signal caused by the variable temperature and humidity. In addition to this, the chosen strain gage arrangement and connection also mutually offset the unavoidable amount of bending caused by the considerable dead weight of the strut and its inclination.
The measuring points were first prepared with Z70 adhesive. The multi-layer system recommended by HBM was used as the covering agent, so that the strain gages were first covered with PU120 polyurethane paint and then with SG250 silicone rubber. ABM75 covering tape provided extra covering for the measuring point once curing was complete. This tape comprises 0.05 mm thick aluminum foil with an underlying 3 mm thick layer of kneading compound. It provides a highly effective barrier against moisture penetration.
Fig. 9: Test device for calibrating the tensioning elements
A measuring and tensioning element like this is shown in Figure 8. A model with identical construction to the measuring and tensioning element was calibrated experimentally in the laboratory. The load frame shown in Fig. 9 was made so that part of the tensioning device, with a strain gage measuring point already attached, could be subjected to a real load by a hydraulic system. A force transducer was inserted so that the force could be measured. From the measured values obtained at different load levels, it was possible to establish a sensitivity to determine the strut force from the bridge output voltage. The measured values of the various load levels show that there is a very good approximation to a linear correlation between these quantities across the entire measuring range.
3.4 Accompanying geodetic measurements
During the redirection, the relative measurements of the online measurement system were supplemented by geodetic methods of measurement. Although the accuracy of these measurements must be given a far lower rating, they do have the advantage of an absolute deformation reference to the solid exterior masonry. This allows the electronic measurements to be checked independently. The following quantities were recorded
- The altitude of the truss frame posts in relation to the exterior masonry
- The altitude of the main runner of the third floor ceiling in relation to the exterior masonry
- The deflection curves of the ceilings at selected points on the second and third floors
The fixed installation of selected geodetic measuring points also allowed changes in deformation to be recorded later on.
3.5 Measured value processing and visualization
The need to visualize the measurement signals on site posed a particular problem. Data from 32 electronic measuring points had to be prepared and summarized to provide visual clarity, so that it was possible for the planner to quickly identify and assess the load and deformation states in the supporting structure at any time during force redirection. The planner also needed to immediately detect when any prescribed limit values were exceeded.
Electrical measured values had to be converted into their corresponding physical quantities in real time and different results were to be determined, such as force totals derived from calculations. A relevant visualization concept was identified in advance of measurement, agreed with the competent supporting structure planner and then implemented.
HBM’s multi-channel Spider 8 amplifier and its associated catman® 5 software were the heart of measured value processing and realization. The combination of catman® and the hardware configuration made it possible to achieve measurement acquisition, data storage and real-time realization of results. A data rate of 1 Hz was considered adequate.
Visualization took place on several monitors, and was simultaneously projected onto screens. Fig. 10 shows the layout of a typical screenshot. The measured press forces and the prescribed limit loads are in the center, and the color of the bars changes if a limit value is exceeded. The associated displacements of the truss frame posts can be seen above and below, and their extent, graded in multiple colors, can be seen at a glance. There are additional calculated values on the right-hand edge of the screen.
Fig. 10:On-site realization of measured values in real time; a Catman online document, projected onto a projector screen
4. Continuous monitoring
The load state of the truss frame struts will now being monitored continuously for a period of two years following the successful force redirection. Data is being acquired and stored by electronically, with a measurement cycle of two hours. The measurement data are regularly transferred and evaluated by remote data transmission, so that they are readily available for assessing the status of the supporting structure.
Seasonal climatic influences are a very important factor in the assessment of the measurement data from long-duration measurements, as changes in the measured value caused by temperature in particular can have a considerable effect on the measurement result. To enable influences of this type to be correctly assessed, both temperatures and relative humidity are also recorded in the course of continuous monitoring.
The measurements that are currently running have already shown that it takes a certain amount of time to establish a stable load state. After a year of continuous monitoring, the tensioning elements were adjusted as defined by the project planner to correct the truss frames loads. It was felt that a complete year was needed following the force redirection to determine the existing stable load state.
