Due to creep characteristics of wood, longterm loading can cause a significant stress loss of steel bars in reinforced glulam beams and high longterm deflection of the beam midspan. In this study, 15 glulam beams were subjected to a 90day longterm loading test, and the effects of longterm loading value, reinforcement ratio and prestress level on the stress of steel bars, midspan longterm deflection, and other parameters were compared and analyzed. The main conclusions drawn from this study were that the longterm deflection of the reinforced glulam beams accounted for 22.5%, 20.6%, and 18.2% of the total deflection respectively when the loading value was 20%, 30%, and 40% of the estimated ultimate load under the longterm loading. The higher the loading level was, the smaller the proportion of the longterm deflection in the total deflection was. Compared with ordinary glulam beams, the longterm deflection of the reinforced glulam beam was even smaller. Under the condition of the constant loading level, the total stress value of the steel bars decreased by 17.5%, 13.6%, and 9.1%, and the proportion of the longterm deflection of the beam midspan in the total deflection was 26.9%, 24.2%, and 20.6% respectively when the reinforcement ratio was 2.05%, 2.68 %, and 3.39%. With the increase of the reinforcement ratio, the stress loss of the steel bars decreased, and the proportion of the longterm deflection decreased as well. When other conditions remained constant and the prestress level of the steel bars was 0 MPa, 30 MPa, and 60 MPa, the total stress value of the steel bars decreased by 9.1%, 9.4%, and 10.2%, respectively, and the proportion of the longterm deflection in the total deflection was 20.6%, 26.1%, and 64.9%, respectively. With the increase of the prestress value, the stress loss of the steel bars increased, and the proportion of the longterm deflection increased as well.
Wooden houses have always been a popular architectural form with the characteristics of safety and habitability. In the context that green buildings are being vigorously promoted around the globe, wooden structures have ushered in new development opportunities due to their merits of low carbon footprint, environmentallyfriendly, and sustainable development [
In recent years, various researchers mainly focused on aspects such as the creep characteristics of wood [
In this study, a 90day longterm loading test was conducted on 15 beams that were divided into 4 groups to explore the impact of loading level, reinforcement ratio, and prestress on the longterm bending performance of the reinforced glulam beams from two perspectives of the total stress variation of the steel bars and the maximum deflection of the beam midspan. Based on the test results, several points that should be paid attention to when designing the reinforced glulam beam are proposed, which also provides the insights for future research.
According to the relevant standards [
According to beam type, loading level, reinforcement ratio, and total prestress level, the specimens were divided into A, B, C, and D groups respectively. Loading level was the percentage of longterm loading value to estimated ultimate load obtained by the finished shortterm loading experiment [
Group  Number of the component  Loading level  Steel bar diameter 
Reinforcement ratio 
Prestress value /MPa  Longterm loading value 
Service load 


A  L_{A1}  20%    0  0  4.66  6.99  
L_{A2}  30%    0  0  6.99  6.99  
L_{A3}  40%    0  0  9.32  6.99  
B  PL_{B1}  20%  18  3.39  0  6.18  9.12  
PL_{B2}  30%  18  3.39  0  9.12  9.12  
PL_{B3}  40%  18  3.39  0  12.36  9.12  
C  PL_{C11}  30%  18  3.39  0  9.12  9.12  
PL_{C12}  30%  18  3.39  0  9.12  9.12  
PL_{C21}  30%  16  2.68  0  9.09  9.09  
PL_{C22}  30%  16  2.68  0  9.09  9.09  
PL_{C31}  30%  14  2.05  0  9.00  9.00  
PL_{C32}  30%  14  2.05  0  9.00  9.00  
D  YL_{D11}  30%  18  3.39  30  9.18  9.18  
YL_{D12}  30%  18  3.39  30  9.18  9.18  
YL_{D21}  30%  18  3.39  60  10.05  10.05  
YL_{D22}  30%  18  3.39  60  10.05  10.05 
Note: PL_{B2} and PL_{C11} in the table are the same beams, but since they are in different groups, they are labeled differently. Subscript C11 and C12 mean identical components; C11, C21, and C31 represent components with different reinforcement ratio, respectively. Subscript D11 and D12 mean identical components; D11 and D21 represent components with different prestress value, respectively.
Due to the large size and the large number of beams for the longterm experiment, a set of longterm loading devices was designed by the research team for the experiment, as shown in
A twopoint symmetrical loading method was adopted in the test. According to the relevant standards [
The measurement in this test included the displacement value of the midspan and the support brackets of the beam, strain of the glulam beams, and strain of the steel bars. To collect the displacement value at the midspan and both ends of the beam, three displacement meters were arranged on these three positions. To collect the strain value of the wood, six strain gauges were uniformly pasted along the side surface at the three equal points on both sides of the glulam part of the test beam. It was worth noting that the defect would be inside the glulam beam. For defective parts, such as microcracks, the strain measured by the strain gauges would be too large under the same deformation. The length of the strain gauges was 100 mm so that the measured strain could be within a certain range, which could largely eliminate the influence of the internal defects on strain measurement results. Moreover, two strain gauges were attached to the top and bottom surfaces of the beam midspan, respectively. As for the reinforced glulam beam, one strain gauge was attached to the middle position of two steel bars at the groove of the beam bottom, as shown in
The prestress was applied by tightening the end nuts (pointed by the arrow) and transferred to the glulam beam through the anchor plates, and then the prestressed glulam beam was formed with compression at the bottom and tension at the top, as shown in
According to the previous research results [
For subsequent analysis, the glulam beams were observed and photographed every 30 days, and the phenomenon in the whole creep process were recorded from the beginning to the end of the experiment. A state diagram of the glulam beams at different times in the longterm experiment was shown in
As could be seen from
In the longterm loading test, the stress level of the steel bars was an important indicator for the mechanical performance of the reinforced glulam beam. In this test, the stress variation of the steel bars was caused by two factors, one was the relaxation of the steel bars, and the other was the creep characteristic of the glulam. Previous studies showed that the relaxation of the steel bars was not significant anymore when the initial stress of the prestressed steel bars was less than 0.5 times value of the ultimate tensile strength
For the beams in group B, C, and D, the variation of the total stress of the steel bars (namely the stress sum of all the steel bars in a beam, as for the reinforced glulam beams in group C and D, the total stress means the average value of the total stress of two identical beams) with time in the whole process from the initial loading to the end of loading was shown in
It was known from
To clarify the change rate of the steel bars stress value, the initial stress of the steel bar was set to 0, The variation of steel bar stress value during the experiment was called relative stress value and the timevarying curve of the relative stress of the steel bar was shown in
In the longterm loading test, observing and recording the development of midspan deflection had great significance in studying the creep performance of the glulam beam. For each batch of the beams, the data were sorted to obtain the longterm deflection development rule of the glulam beams, and the timemidspan deflection relationship curve was drawn as shown in
To facilitate comparison, the downward deflection was specified as positive, and the upward deflection was specified as negative. For prestressed beams, antiarch would appear, so the inverted arch value of midspan was negative. The midspan deflection which generated instantaneously after the applying of the external load was identified as shortterm deflection. After the 90day longterm loading, the extra beam deflection caused by the creep of glulam was longterm deflection, and the sum of the shortterm deflection and the longterm deflection was the total deflection. The inverted arch value, the shortterm deflection, the longterm deflection, and the total deflection of the beams of A, B, C, and D test groups were shown in
Beam number  Inverted arch value (mm)  Shortterm deflection (mm)  Longterm deflection (mm)  Total deflection (mm) 

L_{A1}    5.64  2.41  8.05 
L_{A2}    8.07  3.19  11.26 
L_{A3}    15.44  5.47  20.91 
PL_{B1}    6.24  1.81  8.05 
PL_{B2}    9.64  2.50  12.14 
PL_{B3}    18.87  4.19  23.06 
PL_{C1}    9.64  2.50  12.14 
PL_{C2}    10.84  3.47  14.31 
PL_{C3}    11.77  4.34  16.11 
YL_{D1}  −5.50  6.47  2.28  8.75 
YL_{D2}  −9.79  2.15  3.97  6.12 
As could be seen from
In order to analyze the longterm creep rate of the beam more intuitively, the shortterm deflection was ignored, and then the timevarying curve of the longterm deflection in the glulam beam was shown in
As shown in
For further study of the creep behavior of the reinforced glulam beam, this paper took the average value of the strain gauges at the loading point on both sides of the glulam beam, the strain gauge layout was shown in
The longterm test in this paper lasted for 90 days. However, the design service life of the general structure was 50 years in the building structure. Combined with the previous research experience [
A powerlaw model for the fitting of experimental data and the prediction of the creep deformation of the glulam beams in the longterm experiment was used in this test. The timemidspan deflection formula could be derived from the powerlaw model expressed by
In the formula:
t—the time of creep deformation of the glulam beam (d);
a, b, c—experiment parameters, derived from the fitting curve;
Through the custom function of the origin software, the powerlaw model formula was used for the fitting analysis of the creep data of the glulam beams in the longterm experiment. In order to control the length of the article, the creep fitting curve of the L_{A1} beam was shown in
Beam number  Parameters  R^{2}  

a  b  c  
L_{A}  7.87  0.82  0.32  0.99 
PL_{B}  8.75  0.86  0.31  0.98 
PL_{C}  11.70  1.07  0.32  0.99 
YL_{D}  1.91  1.11  0.30  0.98 
Note: L_{A} represents the ordinary glulam beam; PL_{B}, PL_{C} represents the reinforced glulam beam with reinforcement ratio of 3.39 and 2.68, respectively; YL_{D} represents prestressed glulam beam.
In this paper, the creep deformation coefficient
In this formula:
The specific value of the creep deformation coefficient calculated by
Beam number  Longterm deformation of 
Shortterm deflection/mm  Elastic deflection/mm  Creep deformation coefficient 

L_{A}  26.99  8.07  8.07  2.34 
PL_{B}  26.35  9.64  9.64  1.73 
PL_{C}  36.32  11.77  11.77  2.09 
YL_{D}  23.43  2.15  11.94  1.78 
According to the research content in this paper,
The longterm deflection of flexural members under normal service conditions included the elastic deflection caused by loading, and the creep deflection caused by the creep of wood under continuous load. The elastic deflection
In the formula:
For the prestressed beams, the application of prestress would make the beams produce a corresponding reverse arch. The deflection change of the prestressed beams during the application of prestress and the whole loading process was shown in
The calculation of longterm deflection of the prestressed beam was shown in
In the formula:
When the loading level was 20%, 30%, and 40% of the estimated ultimate load, for the beam midspan of ordinary glulam, the proportion of the longterm deflection in total deflection was 29.9%, 28.3%, and 26.2%, respectively; while for the beam midspan of glulam with reinforced steel bars, this proportion was 22.5%, 20.6%, and 18.2%, respectively. With the loading level increased, the proportion of the longterm deflection due to wood creep effects in the total deflection decreased.
In the case of unchanged loading level, the total stress value of the steel bars in the reinforced glulam beams decreased by 17.5%, 13.6%, and 9.1%, respectively when the reinforcement ratio was 2.05%, 2.68%, and 3.39%. The higher the reinforcement ratio was, the less the stress loss of the steel bars was. For the reinforced glulam beams, the proportion of the longterm deflection in the total deflection was 26.9%, 24.2%, and 20.6% respectively. With the increase of the reinforcement ratio, the proportion of the longterm deflection due to wood creep effects in the total deflection decreased.
In the case of the same loading level and same reinforcement ratio, after subjected to longterm loading, the total stress value of the steel bars decreased by 9.1%, 9.4%, and 10.2%, respectively, and the proportion of the longterm deflection in the total deflection was 20.6%, 26.1%, and 64.9% respectively when the prestress of the steel bars was 0 MPa, 30 MPa, and 60 MPa. With the increase of the prestress value, the stress loss of the steel bars increased, and the proportion of the longterm deflection in the total deflection increased as well.
The author thanks Prof. Hongliang Zuo, Prof. Guodong Li, and Prof. Jianmin Zhang for participating in the wood and other experimental equipment purchase for this study.