Professional Documents
Culture Documents
This study evaluates the failure strength and failure modes of two types of T-joint used in a composite bogie frame under a bending
load. The average failure load and displacement of the adhesive-only joint were 4.1 kN and 0.99 mm. The cracks propagated through
the bond layer and followed a mode I propagation mode. In contrast, the joint with the adhesive bonding and skin had an average
failure load of 68 kN and a displacement of 8.1 mm. It also showed a mode I propagation trend. The strain distribution for the whole
region of the two joints was evaluated with FE analysis. A comparison of experiment results and the FE analysis results revealed that
the error was in range of 5.33% - 42.6% for the adhesive-only joint and 1.07% and 17.3% for the joint with the adhesive bonding
and skin. The maximum stress concentration locations showed good agreement between the experiment and FE analysis.
Fig. 1 Composite bogie frame and T-joints: (a) adhesive-only joint, (b)
joint with the adhesive and skin
Fig. 3 Test setup for the T-joint bending test: (a) adhesive-only joint,
(b) joint with the adhesive and skin
Fig. 2 Manufacturing process for the T-joints: (a) cross beam part lay-
up, (b) vacuum-packing, (c) side beam bonding with cross beam, (d) Fig. 4 Locations of strain gauges: (a) adhesive-only joint, (b) joint with
putting strain gauges on the joint with the adhesive and skin the adhesive and skin
The first one was a T-joint in which the cross beam and side beam were were laid up on the assembled part to a target thickness of 15 mm
connected using only an adhesive bonding method (Fig. 1(a)). The as shown in Fig. 2(d).
second one was a T-joint in which the cross beam and side beam were
assembled using adhesive bonding and a skin (Fig. 1(b)). 2.2 Test setup
To fabricate the adhesive-only T-joint specimens, a steel mould with In this study, the failure strength and behavior for the joint
cross sectional dimensions of 140 mmx-140 mm was manufactured. region was evaluated under a bending load. To apply the bending
Then, the GEP224 glass/epoxy prepregs were laid up to the target load to the exact loading point of the T-joints, a steel beam was
thickness of 15 mm on the mould (Fig. 2(a)). After the lay-up, it was fastened with the T-joint using bolts as shown in Fig. 3. The T-joints
sealed and pressured using a vacuum bag (Fig. 2(b)) and then it was were fixed on the fixing jig using mechanical fastening. To apply
cured in an autoclave. After the curing, a hollow GFRP rectangular the bending load a 5-ton capacity hydraulic actuator (MTS, USA)
tube was made and cut appropriately. Finally, it was bonded with a side for the adhesive-only joint and a 25-ton capacity hydraulic actuator
beam part made of the same prepreg using EPIKOTETM MGS BPR (MTS, USA) for the joint with the adhesive bonding and skin were
135G epoxy resin (Hexion, Germany). used. The deflection at the end of the T-joint was measured using
To make the joints with the adhesive bonding and skin, two a LVDT (Tokyo Sokki, Japan), which was located in the bottom of
tubes of the same size with inner cross sectional dimensions of the joint. The loading rate of the test was 0.5 mm/sec.
140 mm × 140 mm and a thickness of 15 mm were manufactured Fig. 4 shows locations of the strain gauges bonded on the two types
and bonded to each other using EPIKOTETM MGS BPR 135G of joints. A total of 11 gauges (TML, Japan) were used. Four single
epoxy resin (Fig. 2(c)). Then, the GEP224 glass/epoxy prepregs gauges symbolized by S and seven biaxial gauges symbolized by C.
INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 14, No. 1 JANUARY 2013 / 131
Fig. 7 Fractured surface of adhesive-only joint: (a)-(b) specimen 1, (c)-(d) Fig. 9 Load-displacement curve of the joint with adhesive bonding and
specimen 2 skin
Table 1 Strain values comparison of tests and FE analysis results for the adhesive-only joint
Test1 Test2 Average strain
Gauge Name FE (µε) Error (%)
Strain (µε) Status of gauge Strain (µε) Status of gauge (µε)
C21 3607.8 2033.5 Failure before max. load 3800 No failure 3800.0 5.33
C22 5222 2192.6 Failure before max. load 2738.6 Failure before max. load - -
C26 -2317.2 -3104 No failure -2343.1 No failure -2723.6 17.5
C27 -1944.5 -3061 No failure -2485 No failure -2773.0 42.6
S12 764 412.7 Failure before max. load 971.6 No failure 971.6 27.2
Table 4 Strain values comparison of tests and FE analysis results for the joint with the adhesive bonding and skin
Test1 Test2 Average
Gauge Name FE (µε) Error (%)
Strain (µε) Status of gauge Strain (µε) Status of gauge strain (µε)
C21 3132.9 3099.5 No failure 2590.2 Failure before max. load 3099.5 1.07
C22 7463.0 7788.3 No failure 5189.3 Failure before max. load 7788.3 4.36
C26 -7480.0 -8111.2 No failure -8217.6 No failure -8164.4 9.15
C27 -3132.9 -3960.0 No failure -3390.2 No failure -3675.1 17.3
4. Conclusions
5. Guenther, C., Leo, R., and Wackerle, P., “New Technologies for
Rail Vehicle Bogies and Car Body Substructures,” MRS, Europe,
pp. 89-95, 1985.
7. Maurin, L., Boussoir, J., Rougeault, S., Bugaud, M., Ferdinand, P.,
Landrot, A. G., Grunevald, Y.-H., and Chauvin, T., “FBG-based
Smart Composite Bogies for Railway Applications,” Proc. of 15th
Optical Fiber Sensors Conference Technical Digest, pp. 91-94, 2002.