CASE STUDY

Some ideas on the inspection of aluminium mountain bicycle frames Submitted 19.02.14 Accepted 14.03.14

M Giorgetti, M Ruch, J Fava, M Tacchia and E N Olivar Godaz

Different non-destructive tests carried out on Al 7005 mountain bicycle frames are described. The work was originally undertaken in order to prepare an eddy current technique for the NDT of high-end bicycles at different stages of construction and/or trading. The tested frames had been disposed of after six months’ service (the distance travelled is not known) as a consequence of cracking of the head tube near the insertions for the stem and fork. Both had serial numbers and one of them had EN 14766 marking as well. PC-based eddy current equipment was used, with planar and pancake coils mounted on specially-tailored guides, and a calibration piece with artificial defects. A visual test prior to the eddy current testing (ET) showed that one frame had a crack in an expanded section of the fork-stem insertion, while the other had some cracks in a heat-affected zone of the seat tube and in the handlebarstem insertion, where the head tube exhibits a diameter change, tapering and two welds. X-ray fluoroscopy was also performed in order to study the welds. The fluoroscopic images and the eddy current indications from all the coils and defects are presented. Keywords: mountain bicycles, eddy current testing, aluminium, flat coils, fluoroscopy.

1. Introduction Aluminium has largely replaced steel for mountain bicycle frames because of its excellent strength-to-weight ratio, which is about 33% that of steel, and because larger diameter, taperwall aluminium tubes can be used without weight penalty. Aluminium can be used to build lightweight bicycle frames with highly-tuned stiffness characteristics using the hydroforming process. In 2007, a mountain bicycle frame with a fatigue crack in the upper part of the head tube was tested at our laboratory[1]. This crack had grown after a short service time and, following complaints from the customer and seller, the importer replaced the frame and gave a guarantee on the new one. Both frames were made of Al 7005 alloy and the new one has now been in service for six years. Due to the rather large number of aluminium mountain bicycle frames being removed from service as a result of early fatigue cracking in the head tube, a project looking at their eddy current inspection Matías Giorgetti, Marta Ruch, Javier Fava, Mauricio Tacchia and Emilio N Olivar Godaz are with Ensayos no Destructivos y Estructurales, Comisión Nacional de Energía Atómica, San Martin, B1650KNA, Buenos Aires, Argentina. Tel: +54 11 6772 7739; Fax: +54 11 6772 7233; Email: [email protected] / [email protected] Matías Giorgetti is also with the Instituto de Tecnología Jorge Sabato, Comisión Nacional de Energía Atómica, San Martin, B1650KNA, Buenos Aires, Argentina. Javier Fava is also with the Universidad Tecnológica Nacional, Facultad Regional Haedo, Haedo, B1706EAH, Buenos Aires, Argentina.

Insight Vol 56 No 5 May 2014



was carried out at our laboratory. Two more frames made of Al 7005 alloy, which had been disposed of due to cracking in the head tube[2,3], have been tested. A study was carried out of some of the standards applicable to these bicycles[4]. Standards require that different components and/or systems be submitted to destructive tests for fatigue, static resistance, impact and flexural fatigue, among others. Braking tests in dry and wet conditions must also be performed. Other authors have also studied bicycle frame failure. In[5], failure analysis of broken aluminium frames is presented; the techniques used included metallography, fatigue analysis and modelling. In other articles, a failure of the suspension system is studied[6], or the optimal torque and mechanical performance of the screws that hold the handlebar stem[7]. There is also research on the optimal frame design concerning materials and geometry[8], as well as a study on techniques and parameters for the welding of alloys of the Al 7000 (Al-Mg-Zn) series[9]. Pulse infrared thermography can be used to assess the structural integrity of carbon fibre-reinforced frames[10]. In the present work, we describe the application of different non-destructive tests to the study of aluminium frames, which could be used for the inspection of high-end bicycles at different stages of manufacturing, selling or use. Visual tests were performed, an ET technique was developed and, for a more complete evaluation, X-ray fluoroscopy images of some of the welds were made[11]. 1.1 Standards

The EN 14766[4] standard specifies the European Union safety and performance requirements for the design, assembly and testing of bicycles and sub-assemblies intended for off-road rough-terrain use. Other European standards refer to different kinds of bicycles: q EN 14764: City and trekking bicycles. Safety requirements and test methods q EN 14765: Bicycles for young children. Safety requirements and test methods q EN 14781: Racing bicycles. Safety requirements and test methods. Argentina has two such standards: q IRAM 40020: City and trekking bicycles. Safety requirements and test methods q NM 301: Bicycles for young children. Safety requirements and test methods. Some of the properties to be tested are related to safety, performance and riders’ comfort (which include screws and fasteners, brake tests on dry and wet soils and the absence of sharp edges), as well as the mechanical properties of critical parts, to be assessed through strength, fatigue, static and impact tests. Dye penetrant inspection is suggested for the detection of cracking during fatigue tests. The EN 14766[4] standard specifies the test methods and the safety and performance conditions to be considered during the

1

design, construction and testing of mountain bikes. It is clearly stated that: ‘The aim of EN 14766 has been to ensure that bicycles manufactured in compliance with it will be as safe as is practically possible. The tests have been designed to ensure the strength and durability of individual parts as well as of the bicycle as a whole, demanding high-quality throughout and consideration of safety aspects from the design stage onwards’. Furthermore, in Chapter 5, ‘Manufacturer’s instructions’, it is required that ‘each bicycle shall be provided with a set of instructions in the language of the country to which the bicycle will be supplied, containing information and recommendations on different maintenance, safety and performance issues’. In particular, manufacturers are required to include ‘an advisory note to draw the attention of the rider to possible damage due to intensive use and to recommend periodic inspections of the frame, fork and suspension joints (if any)’. In addition, all bicycles satisfying EN 14766 specifications shall be ‘visibly and permanently marked with a successive frame number at a readily visible location, such as near the pedal-crank, the seatpillar or the handlebar, and visibly and durably marked with the name of the manufacturer or the manufacturer’s representative and the number of this European standard’. Within this framework, and just to mention only two manufacturers, on the Cannondale official website there is a section devoted to the user manuals of their bicycles[12], recommending different inspection and maintenance procedures to enhance safety and prevent catastrophic failure. This manufacturer says it follows ASTM F-2043 and European Community standards. This kind of information is also presented on Scott Sports’ official website[13].

Figure 3 shows details of the cracks in the head tube. Crack 1 is the longest, while cracks 2 and 3 are shorter and their roots lay in the heat-affected zone (HAZ) of the weld joining the head tube to the top tube of the frame.

Figure 2. Frame B: (a) head tube, handlebar stem insertion, top tube on the right; (b) seat tube, top tube on the right, seat stays on the left

1.2 Materials

In the present work, the inspection of two Al 7005 (Al-Zn-Mg) frames, ‘A’ and ‘B’, is described. Frame A is made of hydroformed Al-7005 O8[14,15] tubing and has the EN 14766 marking. The section of the tubes making the front triangle is not circular. Frame B is also made of hydroformed Al 7005 tubing. It has no EN 14766 marking. Results from eddy current testing of these frames, with MAD 8D equipment from Eddy Current Technology Inc and a handheld pancake coil from Zetec, were presented in[2]. The design and construction of special probes and positioning devices for the tests were presented in[3], as well as the indications for frame B obtained with planar coils. In the following sections, the probes will be shown and the results of the different methods and techniques will be discussed[11].

2. Non-destructive testing 2.1 Visual testing

Both frames show cracking in the expanded ends of the head tube, where fork and handlebar stems are inserted. The photographs in Figure 1 show the cracks in frame A. The welds of frame A have been polished prior to painting. Some cracks can be observed in the handlebar-stem insertion in the head tube of frame B (Figure 2(a)) and in a region in the seat tube, between the welds to the top tube and the seat stays (Figure 2(b)). In the head tube, a conical expanded region and welds can be observed.

Figure 1. Cracks in the head tube of frame A: (a) fork-stem insertion; (b) handlebar-stem insertion, top tube on the right

2

Figure 3. Cracks in the head tube of frame B. On the right, the top tube can be observed

2.2 Eddy current testing

The eddy current tests were carried out with a MAD 8D®. The first tests were carried out with a hand-held Zetec probe[1,2]. Later, a probe positioning system for the head tube of frame B was made[3,11]; pancake and planar coils[17], as shown in Figure 4, were used. The probe positioning system is inserted in the head tube and can be manually rotated, thus allowing the inspection of the tube wall at a constant lift-off. Holders were made for both types of coil. Figure 5 shows some views of the device for the planar coils: (a) inside the head tube; (b) the outer reference coil holder; and (c) the inner inspection coil holder. The device was constructed in acrylic. Pictures of the inspection device, with planar and pancake coils mounted in the head tube, are presented in Figure 6. Good coupling has been achieved. 2.2.1 Indications

The head tube of frame B was inspected with the three available coils, in order to analyse their sensitivity to the different cracks observed in the visual test (Figure 3). Tests with all the coils were carried out at 90, 100, 180 and 200 kHz. Experiments at 360, 400, 720 and 800 kHz were performed with the planar coils. The coils were rotated in an anticlockwise direction, starting to the left of crack 1. Figure 7 shows the indications from a scan of the outer surface with the hand-held pancake probe. Figure 8 shows the indications from a scan of the inner face with the positioning device and a planar coil (top) and pancake coil (bottom). There is a difference in the opening of the planar coil indications in Figure 8(a). The signal from crack 1 is narrow and clean, while that from crack 2, which starts near the weld, and that from crack 3, clearly on the heat-affected zone, are wider and the starting point Insight Vol 56 No 5 May 2014

(balance point) has moved. All the indications from the pancake coil, Figure 8(b), are narrow and clean, indications from cracks 2 and 3 are parallel and a displacement of the starting point of the signals can also be observed. The displacement of the starting point can most likely be associated with the different metallurgical conditions of the heataffected zone with respect to the rest of the head tube.

Figure 4. (a) Planar coil mounted in its holder; (b) pancake coil

Figure 8. ET crack indications from an inner wall scan with the positioning device: (a) planar coil; (b) pancake coil

2.3 Radiography and fluoroscopy Figure 5. Eddy current testing device for planar coils: (a) inside the head tube; (b) holder for the outer reference coil; (c) holder for the inner inspection coil and screw for system rotation[11]

Figure 6. Inspection device within the head tube: (a) planar coil; (b) pancake coil. Crack 1 is clearly visible

Figure 7. ET indications from the cracks in Figure 3, hand-held Zetec coil, 90 kHz: (a) crack 1; (b) crack 3, in the HAZ of the weld

Insight Vol 56 No 5 May 2014



X-ray images of both frames were made at the industrial radiography laboratory of ENDE, in order to study the quality of the welds in some tube joints. The first two (Figures 9 and 10) are from the front section of frame A (top tube on the left), while the others (Figures 11 and 12) correspond to frame B (top tube on the right). The radiograph in Figure 9 was made with Andrex 200 kV equipment and a D7 plate. The other three images were obtained using X-ray fluoroscopy, a technique which does not use plates but directly processes the digital images. The test conditions are presented in Table 1. The fluoroscopic images were processed with software for the analysis of computed radiography, currently in use in the lab. The plate in Figure 9 was digitised with a General Electric FS50 scanner for industrial radiography. Figure 10 shows the fluoroscopic image of the same area. Variations in the radiographic density can be observed in the insertion of the top tube. Figure 11, front section of frame B, shows the inspection device inside the head tube. There are differences in the radiographic density. The image in Figure 11(a) had normal processing, while that in Figure 11(b) was digitally filtered. Crack 1 (Figure 3) is more clearly seen in the latter than in the former. On the other hand, the welds are more clearly seen in the ‘normal’ image. In the back section of frame B, Figure 12, there is a crack in the seat tube, which most likely started in the HAZ of the joint with the top tube and grew along the edge of the weld up to the HAZ of the seat stays. The central portion of this crack was also observed in the visual inspection. The indication in the lower part of the HAZ

3

Figure 9. Front section of frame A. Technique: plate radiography. Digitised for presentation

Figure 11. Head tube of frame B with the ET positioning device. Digital fluoroscopy, processed with software for computed X-ray analysis: (a) normal; (b) filtered. On the right-hand end there is a crack, which appears with higher contrast in the filtered image

Figure 10. Front section of frame A. Technique: fluoroscopy. Digital image processed for presentation

could be another crack or the projection of the weld on the opposite wall of the tube. A lack of contact between the seat tube and the stays can be observed. These indications are seen more clearly in the filtered image. The welds can be appreciated and crack indications can be detected in the images made with both radiographic techniques. In order to make full radiographic evaluation of these regions, images under different angles should be made, because of the complex geometry of the welds in the frames. On the other hand, welding specifications for the different unions would be necessary to define whether some indications are really defects, such as the non-contact zone between the stays and the seat tube in Figure 12.

3. Conclusions The quality of the reported signals shows that it is possible to inspect bicycle frames made of aluminium alloys using eddy currents. The indications with the different probes are similar. The positioning device was tailored to the particular samples studied. It would be better to design a more versatile device, adaptable to different head tubes. X-ray fluoroscopy and image processing techniques give plenty of information on the welds and cracks. EN 14766 recommends dye penetrant testing to detect the onset of cracking during the destructive tests. The use of ET would allow for the detection of non-wall through cracks, while radiography would enable analysis of the welds. It would be interesting to study the reasons for this early cracking of Al 7005 bicycle frames.

Table 1. Test conditions – plate radiography and digital fluoroscopy X-ray equipment

Technique

Voltage (kV)

Current (mA)

Source-to-sample distance (cm)

Number of shots per image

Frame A

Andrex 200 kV

D7 – Plate – Rx

70

2.5

70

Single plate

Frame A

Philips 160 kV

Fluoroscopy

68

30

30

100

Frame B

Philips 160 kV

Fluoroscopy

70

26

80

100

4

Insight Vol 56 No 5 May 2014

D Torres and P Traverso, ‘Rango de torque seguro y comportamiento mecánico de tornillos de expansión empleados para el ajuste de portamanubrios de bicicletas de uso público’, CONAMET/SAM, 2006. 8. F Dwyer, A Shaw and R Tombarelli, ‘Material and design optimisation for an aluminium bicycle frame’, Worcester Polytechnic Institute, April 2012. 9. G Fu, F Tian and H Wang, ‘Studies on softening of heataffected zone of pulsed-current GMA welded Al-Zn-Mg alloy’, Journal of Materials Processing Technology, 180, pp 216-220, 2006. 10. Newsdesk, ‘Thermal imaging cameras help detect material failures in bicycles’, Insight, Vol 53, No 11, November 2011. 11. M Georgetti, ‘Detección y estudios no destructivos de la fisuración de cuadros de bicicleta’, Tesis de Maestría en Ciencia y Tecnología de Materiales, IT Jorge Sabato, UNSAM, 2013. 12. http://www.cannondale.com/manuals/ 13. http://assets.scott-sports.com.s3.amazonaws.com/manuals/ 14. Registration record of international alloy designations and chemical composition limits for wrought aluminium and wrought aluminium alloys, The Aluminum Association. 15. M Glassel, ‘Tratamientos térmicos de aleaciones de aluminio’, Aluvi, May 2001. 16. Easton Technology Reports, R/D 1, Aluminium alloys, R/D 11, Tube shapes. 17. J Fava and M Ruch, ‘Design, construction and characterisation of ECT sensors with rectangular planar coils’, Insight, Vol 46, No 5, May 2004.

Figure 12. Frame B. Seat tube (vertical), top tube on the right, stays on the left. Welds. Technique: fluoroscopy, processed with software for the analysis of digital radiography: (a) normal; (b) filtered. A crack indication between the welds, a part where there is apparently no contact between the seat tube and the stay

Acknowledgements

The authors wish to thank the Argentine Atomic Energy Commission and Instituto Sabato for their support of this work, Santiago and Pablo, who supplied the frames and encouraged this work, and Esteban Bisignano, who constructed the probe positioning device. References

1. M Ruch, ‘Estudios no destructivos de fisuras en cuadros de bicicleta’, Informe interno ENDE – CNEA, IN-13-E070-IM/10, 2010. 2. M Ruch and M Giorgetti, ‘Fisuras en cuadros de mountain bike. Ensayos no destructivos’, TALMA 2011, La Plata, Mayo 2011. 3. M Georgetti, M Ruch and J Fava, ‘Inspección de cuadros de mountain bike’, 8º CORENDE, Campana, 2011. 4. UNE-EN-AENOR 14766, ‘Bicicletas de montaña: Requisitos de seguridad – métodos de ensayo’, October 2006. 5. S Cicero, R Lacalle, R Cicero, D Fernández and D Méndez, ‘Analysis of the cracking causes in an aluminium alloy bicycle frame’, Engineering Failure Analysis, 18, pp 36-46, 2011. 6. H Shelton, J Obie Sullivan and K Gall, ‘Analysis of the fatigue failure of a mountain bicycle front shock’, Engineering Failure Analysis, 11, pp 375-386, 2004. 7. D Martinez Krahmer, G Maceira, G Papzuck, H Lorusso,

Insight Vol 56 No 5 May 2014



5