Case Studies – Examples of Field Projects Using Digital Radioscopy
St. Anne’s Church, Prague – Examination of Medieval Roof Framing
St. Anne’s Church in Prague, Czech Republic was built in the 1300s. Currently on the World Monuments Fund’s Watch List as an endangered site, the church has functioned as a warehouse for the past two centuries. The original roof timbers are largely intact and appear to be in excellent condition. Questions about whether the trusses had once been covered to produce a wooden vault led to an opportunity to use digital radioscopy to examine some of the truss members.
A catwalk at the base of the trusses allowed for placing the x-ray source on a tripod close to the truss. The image plate was taped to the back of the timber, as shown in Figure 92. Using a distance from the truss of approximately 60 cm and 10 pulses resulted in the radiograph shown in Figure 93. Of interest was the presence of any Gothic nails that may serve as evidence of boards attached to the timber at one time. Such a nail can be seen in the lower left side of the radiograph.
While at St. Anne’s Church, visual inspection of some of the timbers revealed minor pockets of deterioration due to decay fungi. The radioscopy equipment was configured again as shown in Figure 92. The radiograph (Figure 94) was colorized to accentuate the differences in absorption of x-ray energy by the timber. Converting the image to grey scale still allows for noticeable differences in absorption of x-ray energy. The light-colored area in most of the radiograph ￼represents sound wood. The approximately round, dark area is where a void could be seen in the timber. Although the loss of wood could not be determined visually, it was possible to estimate the extent of the void by measuring the dark area on the radiograph. Using the technique described in the section on Quantification of Decay, the maximum void was estimated at 50 percent loss of section.
St. Bernard’s Episcopal Church – Hidden Fasteners
The goal of the radioscopy in this study was to identify any evidence of the types of connections within the wood in the connections of the roof trusses, such as iron straps, iron bars, split ring connections, or mortise and tenon joints. Scaffolding allowed for access to the connection locations shown in Figure 95. Location A, at the peak of the truss where the post is connected was evaluated. Location B1, where the tie connects to the diagonal truss chord (which acts somewhat like a principal rafter) was evaluated. Location B2 is the connection between the toe and post at the peak of the arch. Location C could not be evaluated because of the lack of space to set up the imager at the base of the truss chord. Location C is where the majority of the resistance drilling was conducted.
Figure 96 shows the placement of the x-ray source to generate an image through the cross section at Location B1. Figure 97 shows the placement of the imager for this location. There are two metal straps (front and back) secured by a bolt at one end of the straps through the tie and the other end secured by a bolt through the diagonal truss chord (Figure 97). The source position was offset to see both the front and back straps for purposes of identifying any hidden fastener located between the straps.
The radiograph corresponding to this setup is shown in Figure 99. The image is as seen from the position of the x-ray source and is essentially the left side of the view seen in Figure 98. In the radiograph, the two metal straps are visible, as is metal conduit at the top of the image. The conduit is on the back of the truss tie and is, therefore, not visible in Figure 98 but can be seen in Figure 97. The diagonal line is the edge of the diagonal truss chord. To the right of the truss chord is the truss tie. Nails that secure wood trim to the tie are visible at the bottom of the radiograph. The uneven surface at the bottom of the wood trim is due to charring of the trim.
Note the lack of steel rods or other connectors between the metal straps. This image does not show either of the bolts on the straps because of the position of the source and imager. It is possible to see the grain of the wood in both the diagonal and the tie. There is no mortise and tenon connection between the diagonal and the tie. However, the wood member making up the truss tie does overlap the diagonal truss, as seen (barely) by the cross-hatched pattern of wood grain in the rectangular box in Figure 99.
To attempt to capture any internal fasteners from another perspective, as well as determine the condition of the bolt visible on the surface, the setup was reconfigured as is shown in Figure 100. The imager was placed on top of the tie and the source was positioned to allow x-rays to pass through the tie from the bottom. With this configuration, the bolt should be visible. Figure 101 shows that it is. The bolt is in good condition.
Of interest was what else might be found to support the connection between the tie and the diagonal. No metal fasteners were found in any of the radiographs. The wood member in the truss tie and the diagonal chord simply overlap (like a sandwich). Other than a few nails, the two bolts and metal straps that are visible are the only connection found at this location.
Location B2, where the truss tie and post meet at the apex of the arch below, was the next area of interest (Figure 102). The source was initially placed as shown in Figure 103 to determine whether a mortise and tenon joint connected the tie and the post. The imager was subsequently lowered to determine whether any large metal fasteners secured the arch components to the truss tie.
Figure 104 shows the placement of the imager for one of the tests. The corresponding radiograph for that setup is shown in Figure 105. As was found with Location B1, the metal straps and the two bolts through the straps provide the primary connection between the truss tie and the post. The wood members overlap but only nails secure the apex of the arch to the truss tie.
Location A, at the peak of the truss, was the other location investigated with radioscopy. The visible connection was the same as at Locations B1 and B2 – metal straps secured by two bolts (Figure 106).
The x-ray source was initially placed as shown in Figure 107 in an attempt to view fasteners embedded in either the diagonals or the vertical post of the truss. As can be seen in the accompanying radiograph, the internal detail is the same as was found at Locations B1 and B2 (Figure 108). The area where the straps are located was enhanced to verify that no other metal fasteners (other than nails) were used here. None were found. Reconfiguring the setup allowed for confirmation that bolts pass through the diagonals to secure the straps.
New York State Capitol – Internal Construction and Failure Investigation
A key concern for the Senate Chamber of the New York State Capital in 2005 was the condition of one of the ceiling trusses that exhibited visible deflection. The construction of the truss connection details and positioning of blocking or other components was unknown. Destructive testing was not desirable because of the placement of the historic decorative wood trim over the lower truss chord. Digital radioscopy was proposed as a means to examine the truss at key locations identified by the structural engineer to determine the presence or absence of connections and any possible failure of the wooden truss chords.
The purpose of the investigation was to use digital radioscopy to identify evidence of the types of connections within the lower truss chord of one ceiling truss, such as steel plates, iron straps, iron bars, or bolts. Additionally, a possible failure in one of the truss chords was to be investigated. Scaffolding allowed for direct access to the lower truss chord.
Scaffolding was erected below the truss to allow full access along the length of the lower truss chord. The scaffolding provided a platform for placing the x-ray ￼source in various positions to capture images at key locations along the lower truss chord (Figure 109). Note that the decorative trim has separated due to deflection of the truss. The truss measures approximately 18 inches from front to back face. The source was typically placed 12 to 20 inches from the front face of the decorative wood trim over the truss chord.
Prior to conducting the fieldwork, a mock-up was constructed in the lab to determine the likely number of pulses needed to penetrate the steel side plates, if necessary, and the effect of shooting at an angle on the ability to detect internal features. The mock-up was configured from field information provided by the structural engineer. Based on the lab research, it was known that placement of the imaging plate was critical for attempting to accurately locate structural rods, fasteners, plates and, if present, any potential fractures in the truss chords. Shooting perpendicular to the surface of the truss chord was the preferred setup and was used at most of the locations along the truss. Figure 110 shows the position of the source near purlin 8 to shoot perpendicular to the surface of the truss. The imaging plate for this shot is oriented as shown in Figure 111.
Images were captured systematically along the length of the truss chord between three purlins (the area of interest to the structural engineer) so that the radiographs could be “connected” to continuously show the internal construction of the lower truss chord. The schematic shown in Figure 112 shows both the location and orientation of the imaging plate for each shot. The radiographs were oriented to show the internal view of the truss chord as it appears below the imaging plate (as opposed to being viewed from the source side of the truss).
As shown in the radiographs in Figures 113 through 114, the radiographs successfully identified metal connections of the truss system, such as screws, nails, threaded rods, bolts and steel plates. In addition the sensitivity of the radiographs was sufficient to identify more subtle features such as wood grain in the trusses and decorative carved wooden trim. Key elements were labeled on each radiograph. The important features can be summarized as follows:
- Steel plates were present below where purlins 6 and 8 intersected the truss.
- Vertical threaded rods were visible at purlins 6 and 8.
- No steel plate was present below purlin 7 at truss 3.
- A wood diagonal truss chord was visible at purlin 6.
- A probable fracture (as opposed to a seasoning check in the lumber
due to drying) was visible in images 15, 1, 14, 13 and 12 (proceeding from west to east). The fracture extended at least from west of purlin 8 to west of purlin 6, based on the number of images taken. Subsequent destructive testing verified the extent of the fracture.