Notice: This page was translated automatically. If anything seems unclear, please refer to the German version.

Investigations to determine the load-bearing behaviour of geogrids when bridging earth subsidence

Dr.-Ing. S. Schwerdt, Magdeburg-Stendal University of Applied Sciences

Summary

When designing geosynthetic reinforcements to bridge earth failures, the required anchoring and overlap lengths must be determined in addition to the tensile force. When designing reinforcements using biaxially anisotropic reinforcements, the question arises as to how the load is transferred at the edges of the reinforcement. This question was investigated using burst tests.

The tests showed that the strains in the transverse direction were smaller and those in the longitudinal direction were greater at the edges than in the centre of the geosynthetic sheet. Differences were also found in the failure behaviour and bulging. This led to conclusions that will have consequences for the future design of biaxially anisotropic geosynthetic reinforcements for bridging earth failures.

Introduction

Sinkholes are usually crater-shaped depressions on the ground surface. These can be of natural origin, known as sinkholes, which are caused by dissolution and subsoil erosion processes underground. Sinkholes, pits or shaft collapses occur when old mining facilities are inadequately secured. When constructing transport routes or engineering and civil engineering structures in areas at risk of sinkholes, safety measures must be put in place to eliminate hazards.

In principle, various safety options are possible for this purpose. The most economical solution is often the use of geosynthetic reinforcement. Geosynthetic reinforcements are designed using calculation methods that are country-specific and normative in nature, e.g. BS 8006 (1), or correspond to the state of the art, e.g. EBGEO (2). In addition to the required short-term tensile strength of the geosynthetic, the overlap and anchoring lengths must also be determined.

Due to the manufacturing method and materials used, different load transfer behaviours can be observed for different geosynthetics when bridging earth failures. In principle, a distinction is made between biaxially isotropic, biaxially anisotropic and uniaxially extremely anisotropic load transfer (see Table 2).

In the course of investigations into the load transfer behaviour of biaxially anisotropic Secugrid® geogrids, the load transfer behaviour in the anchoring area was examined in particular.

Safety concepts and design methods

Safety concepts

There are numerous regulations for securing structures in areas at risk of ground subsidence, such as the information sheet on road construction in sinkhole areas (3) or GSbS 2006 (4).

The recommendations for the use of geosynthetics to secure road areas at risk of collapse in old mining and subrosion areas for the Saxony-Anhalt State Construction Authority (GSbS 2006) (4) distinguish between the concept of partial securing and the concept of full securing. Both concepts differ in terms of the durability of the solution, the intended protection period and, ultimately, the effort required. Table 1 shows the main differences between the two concepts.

Table 1: Comparison of the full and partial securing concepts; based on GSbS 2006

It can often be assumed that, from a ratio value between the cover height and the earth failure diameter of H/D ≥ 1, the use of geosynthetics and design according to the partial protection concept will lead to more economical solutions.

Design method

According to EBGEO (2), different geosynthetics can be used for bridging earth failure bridges. These are divided into isotropic, anisotropic and extremely anisotropic materials. The type of reinforcement determines the load transfer model to be used in each case (see Table 2). The design can then be carried out using various calculation methods.

Various design steps must be carried out when designing earth failure bridges. These include:

1. Determination of the geometric boundary conditions
2. Determination of the permissible settlement of the geosynthetic
3. Determination of the vertical stress on the geosynthetic 4. Calculation of the tensile force in the longitudinal (and transverse) direction

5. Determination of the overlap and anchoring lengths

The geometric boundary conditions are defined on the basis of external specifications (e.g. permissible subsidence at the top edge of the carriageway). The determination of the settlement of the geosynthetic, the vertical stress and the calculation of the tensile force are described in the various calculation methods (see Table 2). The EBGEO (2) provides guidance on determining the overlap and anchorage lengths.

Investigations to determine the load transfer behaviour of biaxially anisotropic geosynthetics in the anchoring area

Problem

Figure 1 shows the determination of the required anchoring length in the transverse direction of a geosynthetic reinforcement. The required anchoring length begins outside the area to be secured.

The anchorage length is determined according to Table 11.3 in EBGEO. For biaxially anisotropic geosynthetics, this is calculated in the transverse direction using the following formulas:

Where: Ll – anchoring length in the transverse direction; LA,cmd – length of load transfer in the transverse direction; D – earth failure diameter; E_d – design value of the effects; Ȗ_B – partial safety factor; į_vg,k – characteristic vertical stress; f_sg,k – friction coefficient soil/geosynthetic; factor 2 for two-sided load transfer

According to EBGEO, for biaxially anisotropic geosynthetics, the soil failure diameter D does not need to be added to the anchorage length "… if structural measures ensure that no load transfer occurs in the transverse direction." ¹

_______________________

1 (2), 11.3.2.5

In practice, this means that when designing biaxially anisotropic geosynthetic reinforcements for soil failure bridging, in many cases the soil failure diameter must be added to the anchorage length, as the necessary design measures are not specified in detail. This applies in particular to cases where a landslide occurs under the geosynthetic in the anchoring area and the transverse strength is used for load transfer as planned.

The aim of the investigations presented below was therefore to determine how biaxial anisotropic geosynthetic reinforcements (Secugrid® R6) behave at the edge of the reinforcement sheet when a sinkhole occurs immediately below the end of the sheet (see Figure 2).

Fig. 2: Example of geogrid at the end of the sheet above a ground collapse

The following questions were to be addressed:

1. How great are the elongation and tensile force at the end of a sheet above a sinkhole in the transverse direction of the geosynthetic? 2. Are there differences in the tensile force occurring in the longitudinal direction when the sinkhole occurs at the end of the sheet, compared to a sinkhole in the middle under a geosynthetic layer?

3. Are adjustments to the EBGEO design specifications necessary?

The investigations should initially be limited exclusively to biaxially anisotropic geosynthetics.

Test concept

The questions should be clarified with the aid of burst tests, on the basis of which essential research work relevant to the development of EBGEO was carried out in earlier investigations. For this purpose, a "large burst pressure testing device" was used, with which a sample diameter of 1.0 m can be examined. The load is applied by gradually increasing hydraulic pressure. During each load stage, the strain and bulging are measured. Burst tests make it possible to obtain information about the stress-strain behaviour, burst behaviour, ultimate bearing capacity, behaviour in the overlap area and behaviour at the end of a geosynthetic strip in the earth failure area. Figure 3 shows the burst pressure testing machine used.

Fig. 3: Large burst pressure testing device

Burst tests were carried out on biaxially anisotropic geogrids. The Secugrid® 120/40 R6 geogrid from NAUE GmbH & Co. KG was selected for this purpose. According to the data sheet (7), the following parameters are known for this geogrid:

• Tensile force (nominal value): longitudinal/transverse direction 120/40 kN/m
• Elongation (nominal value): longitudinal/transverse direction ≤7/≤7%

To clarify the task, two partial samples were laid in each case, which only touched each other in the transverse direction without overlapping and did not allow any load transfer ("tests with gap") (see Figure 4). Different sample widths were used.

• Test 1:	25 / 75 cm (transverse direction at the top)
• Test 2:	42 / 58 cm (transverse direction at the top)
• Test 3:	46.5 / 53.5 cm (longitudinal direction at the top)

In addition, a reference test was carried out in which the geogrid was laid over the entire surface.

Fig. 4: Two partial samples of the geogrid in the burst pressure testing device before the start of the test

Test results

Reference test

The following figures show the results of the reference test. Failure occurred at 1.3 bar. Multiple failures of the bars occurred in the longitudinal and transverse directions.

Fig. 5: Strains during the reference test

Fig. 6: Change in height as a function of pressure [bar] during the reference test

Tests with gap

In the tests with sample widths of 75/25 cm (test 1) and 58/42 cm (test 2), the geogrid samples were arranged so that the transverse direction of the geogrid was at the top. In the test with sample widths of 53.5/46.5 cm (test 3), the longitudinal direction was at the top. In the tests in which the bars in the transverse direction were at the top, higher pressures led to partial failure of the connection points between the longitudinal and transverse bars. As a rule, this failure only occurred on the outer longitudinal bars. In test 1, this led to the underlying membrane being pushed out due to the longitudinal bars being displaced from the joint, and no further load application was possible. In test 2, the outer longitudinal bars failed at 0.8 bar. In test 3, no damage to the connections was observed. Here, failure occurred at 1.15 bar due to the bars bursting in the longitudinal direction. In reality, when the Secugrid® geogrid is rolled out, the cross bars are on top, i.e. in the bursting test they would then be located under the longitudinal bars. In this arrangement, therefore, no failure of the connections is to be expected, in particular due to the additional shear resistance that becomes effective through the interaction of the soil with the geogrid.

The following figure shows the measured strains from the tests with a gap. The results of the reference test are included for comparison purposes.

Fig. 7: Strains during the course of the tests with gaps and the reference test

It can be seen that the strains in the longitudinal direction (maximum 5.8% at 1.2 bar) are generally greater than in the reference test (4.3% at 1.2 bar). In contrast, the strains in the transverse direction, at 0.0 to 1.5%, are significantly smaller than in the reference test (5.8%). In two tests, failure occurred when the bars burst in the longitudinal direction at the end of the respective test specimens. In one test, no failure of the bars was observed. In this case, it was not possible to apply any further load because the longitudinal bars had shifted on the membrane.

A gap opened between the two test specimens. The opening widths measured are shown in the following diagram.

Up to approx. 0.6 bar, the samples behaved almost identically. After that, a disproportionate increase in the opening width can be seen in the second test with sample widths of 58/42 cm. In contrast, the increase is almost linear in the remaining tests.

A comparison of the bulges in Figure 9 shows that larger bulges were measured in the tests with gaps than in the reference test.

Fig. 9: Bulging in the tests with gaps and the reference test at 0.8 bar

Results and conclusions

The results of the tests carried out can be summarised as follows:

1. At the end of the track, outside the area to be secured, the load transfer behaviour of biaxially anisotropic Secugrid® geogrids differs from that inside the area to be secured.
2. If there is a possibility of landslides occurring outside the area to be secured, load transfer occurs from the transverse to the longitudinal direction.
3. The strains and thus the tensile forces in the transverse direction are significantly smaller than in the reference test. In contrast, the strains and tensile forces in the longitudinal direction increase.
4. In the tests, failure occurred exclusively in the longitudinal direction.
5. The bulges are larger than in the reference test under the same load.

This leads to the following conclusions:

1. Outside the area to be secured, in our opinion, an additional earth collapse diameter should not be applied when determining the anchorage length L_l according to formula (1).
2. When calculating the anchorage length L_A using formula (2), the factor "2" for load transfer on both sides must not be used. Since a soil collapse cannot be ruled out under the load transfer area, one-sided load transfer must be taken into account in future.
3. This results in an increase in tensile strength in the longitudinal direction. Outside the area to be secured,

this is irrelevant, as no securing is required here. However, we consider the behaviour of the last lane at the edge of the area to be secured to be problematic. In our opinion, investigations are necessary here to determine whether greater tensile strength in the longitudinal direction needs to be verified. This can be done, for example, by means of parameter studies using numerical methods.

Acknowledgement

The investigations were initiated and funded by NAUE GmbH & Co. KG. Magdeburg-Stendal University of Applied Sciences would like to express its gratitude for the funding.

Bibliography

1. BS 8006-1. BSI: British Standard Institution: Code of practice for strengthened/reinforced earths and other fills. London: s.n., 2010.
2. DGGT. EBGEO Recommendations for the design and dimensioning of geosynthetic reinforcements. [Ed.] German Geotechnical Society. Berlin: Verlag W.Ernst und Sohn, 2010.
3. FGSV. Information on road construction in sinkhole areas. s.l.: FGSV-Verlag, 2010.
4. LSBB ST. Recommendations for the use of geosynthetics to secure road sections at risk of failure

road sections in old mining and subrosion areas for the Saxony-Anhalt State Construction Authority. Halle: s.n., 2006.

1. Design of Soil Layer – Geosynthetic systems overlying Voids. Giroud; Bonaparte; Beech. 1, 1990, Geotextiles and Geomembranes, Vol. 9, pp. 11-50.
2. Blivet. Design method for geosynthetics as reinforcement for embankments subjected to localised subsidence. [Book author] Delmas, Gourc and Girard (ed). Proc. 7th ICG. s.l.: Sweets & Zeitlinger, 2002.
3. NAUE GmbH & Co. KG. Data sheet Secugrid® 120/40 R6.