Experimental Study on Reasonable Horizontal Load Value on the Top of Railings of Glass Bridges or Gallery Roads in Scenic Area

With the increase of glass bridges or glass walkways in the scenic area, the safety protection provided by railings cannot be neglected. However, there is a diﬀerence between the demands for a horizontal load on the top of railings in the present design speciﬁcations. Especially, it does not give speciﬁc advice on the horizontal load on the top of railings of glass bridges or glass roads in the scenic area. In this study, experimental research is conducted on railing columns based on the investigation of common railings of glass bridges or glass roads in the scenic area. The type of railing columns includes ﬁnished products and ﬁeld production. The number of railing columns is 12. The displacement and strain results of the railing columns under diﬀerent loads are obtained. The rigidity and strength are analyzed by numerical simulation combined with test results. A reasonable horizontal load on the top of the railings of a glass bridge or glass road is advised. These results provide a reference for the design, inspection, and evaluation of railings of glass bridges or gallery roads in the scenic area.


Introduction
On March 2, 2021, in a public university in El Alto, Bolivia, the corridor railings suddenly broke, causing seven students to fall and one student to be seriously injured. In the Lantern Festival, in 2004, a crush and stampede accident occurred on the Rainbow Bridge in Mihong Park in Miyun District, Beijing, resulting in 37 deaths and 15 injuries [1]. ere have also been many accidents with casualties on campus and shopping malls being caused by collapsed railings [2][3][4]. At present, glass bridge and gallery road railing systems in scenic areas are designed and manufactured according to the standards of railing systems of civil buildings. Tourists will stop to enjoy and take photos in the steep and dangerous places. us, the safety protection of the railing cannot be neglected. It is important to ensure the reliability of railings.
In the existing literature [5][6][7][8][9], different railing systems are verified to satisfy the requirements by the loading test. Fan et al. [10] test the horizontal thrust and vertical pressure of a bridge railing. According to the analysis of the finite element method, the most disadvantageous loading control position of the railing is determined by the design uniform load which is equivalent to the test concentrated load. Zhang and Guo [11] studied the long rail at a bus station by the overall detection method and the cut-off test method. And, the horizontal load of 2.5 kN/m was advised for the railing with two people above. In the existing literature, the study of the horizontal load of railings has not been found. e load of railings is specifically advised in the present specification. On the top of a railing, the horizontal load value is 1.0 kN/m, and the vertical load value is 1.2 kN/m, according to Load Code for the Design of Building Structures (GB 50009-2012) [12]. e horizontal load value is 2.5 kN/m and the vertical load value is 1.2 kN/m e according to Technical Specifications of Urban Pedestrian Overcrossing and Underpass (CJJ 69-95) [13]. In addition, according to Design Specifications for Highway Safety Facilities (JTG D81-2017) [14], the standard horizontal loading value on the top of sidewalk or bicycle lane railings is 0.75 kN/m and the standard vertical load value is 1.0 kN/m. According to the uniform building code [15], the load of railings outside the exit is 292 N/m, and the load of railings at the exit is 730 N/ m. According to Eurocode [16], both the horizontal load and vertical load are 1.0 kN/m. ere is a difference between the horizontal load values in different specifications. en, the maximum value of the horizontal load is advised in Technical Specifications of Urban Pedestrian Overcrossing and Underpass (CJJ 69-95).
For human safety, Hopkins et al. [17] investigated the average pressure of a crowd in a New Year's party and a popular concert in New York, USA. e results showed that people will die in ten minutes in body pressure of 2.1 kN/m and will die in 15 to 30 seconds in a body of 11.68 kN/m. Smith [18] developed an inclined crowd model to quantify the pressure exerted by the crowd on the guard railings. Huang et al. [19] analyzed the pressure bearing capacity of people and individuals by the aforementioned inclined crowd model. e results showed that the safe design load value of human bearing capacity increases from 0.75 to 1.0 kN/m. erefore, the design value of the horizontal load of railings should be limited.
For railings, structural safety and human safety both should be satisfied. It is necessary to study the reasonable horizontal load value of railings of glass bridges or gallery roads. In this study, experimental research was conducted on railing columns. e type of railing columns includes finished products and field production. e displacement and strain results of the railing columns under different loads are obtained.
e rigidity and strength are analyzed by numerical simulation combined with test results. A reasonable horizontal load on the top of the railings of a glass bridge or glass road is advised.

Experiment Details
2.1. Simplified Model of Railings. In a railing system, the railing bars are firmly combined with the post which is fixed on the bottom support. e horizontal distributed load F is applied to the railing bars outwards. And, the load is transferred to the post through bar-post joints. e post is the main bearing component. is study focuses on a single post. e horizontal distributed load F of the railing bars is equivalent to concentrated load P applied to the post, as given in equation (1) and Table 1. e concentrated load: where L is post spacing, calculated as L � 1 m.

Test Methods.
ere are two types of railing posts for glass bridges or gallery roads in scenic areas based on an investigation. Type I is steel pipe railing posts which are rectangular steel tubes with a thickness of 2 mm-6 mm and a width of 60 mm-120 mm, as shown in Figure 1. Generally, the railing posts and the bottom support are connected by fillet welds which are 4-6 mm in length. Type II is finished railing posts, assembled on site by the manufacturer, as shown in Figure 2. Type II are single-piece posts made of stainless steel with a thickness of 5 mm-15 mm and a width of 60 mm-120 mm. e bottom gasket of a post is fixed on the bottom support through bolts or welds.
ere are two differences between the aforementioned two types of railings. First, the cross sections of the two types of railing posts are different. Second, two types of railing posts are connected to the bottom support in different ways. Finished railings are generally fastened by bolts or spot welding. e steel pipe posts are generally connected by full welding.
Two types of railing posts were tested. Material properties of the selected steel pipe post and the finished post are shown in Table 2 and detailed size parameters are shown in Table 3. e steel pipe post is made of Q235 steel, while the finished post is made of 304 stainless steel. Posts are fixed on the bottom support by welding with 6 mm fillet welds. e nodes at the bottom of the railing posts are fixed with the actual condition consistent. A horizontal loading test was conducted on the top of the post. e displacement, stress, and other test phenomena were recorded. e maximum test load is 5 kN/m. As shown in Figure 3, the loading system equipment includes a counterforce frame, a load gauge, a displacement sensor, and a strain gauge. e bottom of the post is fixed on the test platform while the top of the post is welded with two railing bars. Displacement measuring points are arranged on the top of the post. Strain measuring points are arranged on its bottom.

Test Results.
e test results of 12 railing posts are summarized in Figure 4. e maximum load, maximum strain, and test phenomena are recorded in Table 4. Posts 1  To sum up, firstly, the effect of post height on the horizontal load value of the steel pipe post is obvious. With the height increase, both the post rigidity and the horizontal load value decreased. Secondly, the maximum test load of the steel pipe post is bigger compared to that of the finished post. At the same load, the strain of the steel pipe post is smaller compared to that of the finished post. irdly, the stress of support of the finished post is large with the horizontal load applied.

Parametric Analysis.
Method of normalization was adopted to analyze the test results of steel pipe post 1 to 8 to study the relationship between railing load at the top P, post height H, section property W, material strength f y , displacement Δ, and strain ε.
(1) e correlation curve of (FH/W · fy) ∝ Δ/H is shown in Figure 5(a). According to Technical Code for Test and Evaluation of City Bridges (CJJ/T 233-2015) [20], H/120 was taken as the horizontal displacement limit. When Δ/H � 1/120, the value of (FH/W · fy) ranged from 0.15 to 0.31. e correlation curve of (FH/W · fy) ∝ ε is shown in Figure 5(b). When (FH/W · fy)<1, ε < 778(×10 −6 ), the maximum strain of components 2, 4, and 7 exceed 500. e maximum strain of other components was less than 400(×10 −6 ). When the value range of (FH/W · fy) ranged from 0.15 to 0.31, the maximum strain is 330(×10 −6 ). (2) When the displacement limit is H/120, the load curve of the steel pipe post is shown in Figure 6. e load value of post 5 was 0.8 kN/m which was the lowest load among posts. e load value of post 8 was 1.9 kN/m which was the highest load. e average load of posts is around 1.2 kN/m.

Numerical Analysis
e numerical analysis method was adopted to research more sizes of steel pipe posts in this study. e model of Qi et al. [21] was adopted. In the study of Qi et al. [21], a comparative study was conducted on the between element model and overall model applied on the analysis of the middle part of the column. e results showed that the results of the internal force and displacement of the middle part column obtained by the element model and overall model are consistent. e element model was proved to analyze the railing system well.
e finite element analysis model was developed with Midas, as shown in Figure 7(a). e element model included a post and two half-span railings. One railing is at the left of the post and the other is at the right side. e post bottom was designed to be fixed support. For ends of the railings, the displacement of x, y, and z directions of the node was constrained. Connections between the railing and the post were rigid. e displacement in x, y, and z directions of connections was constrained. e upper node of the post was adopted as the main node. A nodal load was applied on the post based on the element model. And, loading conditions were the same as the test conditions. e material parameters were taken according to the results of material property experiments, as shown in Table 2, and the type of finite element is the bar element. Calculated results are shown in Figures 7(b) and 7(c). Figure 8 shows the comparison between numerical analysis results and test results.
Compared to the results obtained by the experiment, calculated results of deformation and displacement of the railing post using the element model are consistent. Especially, the calculated results' displacement values are the same as the results from the experiment. erefore, the element model can be used for numerical simulation of the railing posts of glass bridges or gallery roads in the scenic area.
Post height and section size of the post were analyzed. Based on the investigation, most of the steel pipe posts of    glass bridges and gallery roads in the scenic area were square steel tubes. e height of square steel tubes ranged from 1000 mm to 2000 mm. e section dimension of square steel tubes was (40-100) mm * (2-5) mm. Detailed parameters of the element model are displayed in Table 5. As shown in Table 5, 3 height values and 6 section dimensions of steel pipe posts were adopted, and subsequently, 18 element models were developed for numerical calculation. e results obtained from 18 element models were analyzed by the method of normalization. e results show that (1) e relationship of (FH/W · fy) ∝ Δ/H and (FH/W · fy) ∝ ε is shown in Figure 9. e horizontal displacement limit is H/120. When Δ/H � 1/1201/120, (FH/W · fy) has a value range     from 0.10 to 0.56. e equation (FH/W · fy) can be considered as the ratio between post function S and resistance R. It implied that S/R ranged from 0.10 to 0.56, and equation (2) can be obtained:

Advances in Materials Science and Engineering
(2) When the horizontal displacement limit is H/120, the relationship of element model load F and Δ/H is shown in Figure 10. It can be seen that the load of H1B5, H1B6, H2B6, H1B10, and H1B12 were higher than 5.0 kN/m. e load of H3B12 was 0.1 kN/m which was the minimum value. Load values under Δ/H � 1/120 were summarized in Figure 11. e average load was around 1.3 kN/m.   Note: an element model with a height of 1000 mm and a section size of 50 * 50 * 2 are regarded as H1B1.
(3) Based on the results of the experiment and simulation, 39 sets of data were used to estimate the averaged horizontal load. e estimated value of averaged horizontal load was 1.3 kN/m. According to results of statistical analysis, the confidence interval of the averaged horizontal load with a confidence level of 95% is given:

Conclusion
(1) A simplified test model of railings' post was proposed and load tests were carried out on railing posts of different sizes. e results showed that the maximum load of steel pipe posts was 5 kN/m, and the maximum load of finished posts was 2.5 kN/m. Compared with that of finished posts, the maximum load of steel pipe posts was higher and the strain value of steel pipe posts was smaller. erefore, it is not recommended to choose finished posts as the main stress components in the railing systems of glass bridges and gallery roads in the scenic area. (2) Results calculated by element models have the same tendency as that of railing posts. erefore, it is acceptable to use element models for railing post simulation.
(3) Test results of 8 steel pipe posts and 18 steel pipe element models were numerically analyzed. When the displacement limit of the top of the post was H/ 120, FH/W · fy ranged from 0.10 to 0.56, and ε was smaller than 230(×10 −6 ). erefore, H/120 is considered as the displacement limit of the top of the post to ensure that the bearing capacity of the post satisfied the requirements. (4) With the displacement limit of H/120, the average horizontal load of 8 steel pipe posts was 1.2 kN/m and the averaged horizontal load of 18 steel pipe post element models was 1.3 kN/m. Based on analyzing the two sets of data, the confidence interval of the averaged horizontal load with a confidence level of 95% is <μ > 0.95 � (1.0; 1.6) kN/m. erefore, the reasonable value of the horizontal load on the top of the glass bridge or gallery road railings ranged from 1.0 kN/m to 1.6 kN/m.

Data Availability
e data used to support the findings of the study can be obtained upon request (13691371229@163.com).

Conflicts of Interest
e authors declare that they have no conflicts of interest. Advances in Materials Science and Engineering 9