Development of Sustainable and Innovative Manhole Covers in Fibre-Reinforced Concrete and GFRP Grating

Abstract

In several countries, manhole covers made of steel are being stolen, with significant economic losses for private and public entities, and even causing accidents. In this work, a new manhole cover is developed using fibre-reinforced cementitious (FRC) materials and glass fibre-reinforced polymer (GFRP) gratings. Since the GFRP gratings are immune to corrosion, and FRC is a relatively low-cost material, manhole covers in FRC reinforced with GFRP gratings are durable and not so appealing to be stolen as those made from steel. An experimental program with manhole cover specimens made with two types of FRC and two types of GFRP gratings was executed by investigating the strength, stiffness and post-cracking tensile capacity of the FRCs and the stiffness and flexural capacity of the two GFRP gratings. It was demonstrated that the developed manhole cover concept can be of class A15 up to D400 according to the recommendations of BS EN 124:1994.

1. Introduction

Population growth and the direct impact that potable water supply and the sewage network have on the health and life quality of citizens indicate a very high rate of growth in demand for manhole covers over the coming years. Climatic changes result in the need for the upgrading of existing rainwater networks and for the construction of new ones, which also require manhole covers.

Manhole covers have been commonly produced using steel or reinforced concrete (with conventional steel reinforcements). Steel manhole covers can suffer deterioration by corrosion and local ruptures (Figure 1), and have been stolen for resale as steel scrap [1,2,3] causing serious risks for the transients and residents, with numerous cases of disasters, and significant costs for particulars and to the municipalities [4,5].

Manhole covers produced by RC are susceptible to corrosion of the steel reinforcement, which limits the long-term performance of this type of urban infrastructure. This weakness causes the loss of structural serviceability and safety due to the decrease in the concrete–steel bond strength and cross-section of the reinforcements. In recent years, some initiatives have aimed at developing manhole covers exclusively in fibre-reinforced polymers (FRPs). Besides the higher price, the structural performance of these manhole covers is dependent on the glass transition temperature (Tg) of the polymer component of the composite material, since, in road applications, the temperature can exceed the Tg. In fact, temperatures exceeding 70 Celsius degrees have been recorded in pavements [6]. The glass transition temperature of the polymeric matrix used in the major part of the GFRP applied in construction systems is between 65 and 150 Celsius degrees [7]; therefore, the mechanical properties of these GFRPs are detrimentally affected when subjected to the maximum temperatures possible to be attained in pavements. The long-term inelastic deformability of these composites is also significant, which results in their needing replacement in relatively small periods of time.

Fibre-reinforced cementitious (FRC) materials have been combined with FRP in the development of new construction systems in an attempt at mobilizing the potential of these materials in a complementary perspective towards intended objectives like functionality, sustainability, economic viability and structural and functional performance [8,9,10,11,12,13]. Concrete has a relatively high compressive strength and modulus of elasticity, high durability and resistance to high temperatures, and can be engineered to have rheology capable of filling relatively small spaces and moulds of complex shape, from self-compacting to ultra-fluid. However, when subjected to tension, concrete develops brittle behaviour; therefore, small fibres have been used as an extra constituent of concrete composition to increase its post-cracking tensile capacity and energy dissipation and absorption at a relatively low cost. Fibres have been used to eliminate partially or even completely conventional steel transverse reinforcement in beams [14,15,16] and slabs [17,18,19,20], to increase the flexural capacity of members in combination with a certain percentage of conventional flexural reinforcement (hybrid reinforcement, HFRC) [21,22] or to be the exclusive reinforcement [23]. In the case of HFRC structures, fibres can contribute significantly to increasing the load-carrying capacity during the crack propagation phase and decreasing the crack width and the maximum tensile stress in the flexural reinforcement for the load combinations corresponding to the serviceability limit state (SLS) design verifications [24,25,26]. When used as the unique reinforcement, FRC has been used in structures of high redundant support, like slabs on soil [27,28,29], slope stabilization [30,31,32,33] and tunnel lining [34,35], by taking advantage of the relatively high level of stress redistribution capacity of this type of structure during the crack propagation process.

Due to the susceptibility of steel to corrosion [36], FRP rods have been used in combination with steel rods (hybrid flexural reinforcement, HFR) or as unique flexural reinforcement. In HFR, the steel rods are used to assure the load-carrying capacity of the structural member in case of a fire occurrence, since the properties of FRP rods are detrimentally affected by temperatures exceeding the Tg of the polymer used in their production [37,38]. In the HFR, steel rods are also used to guarantee the aimed level of ductility, since FRP bars have linear-brittle tensile behaviour. Furthermore, FRP rods with a smaller modulus of elasticity and bond performance to the surrounding concrete and steel rods are used in structural applications, which has detrimental consequences for the SLS design verifications [39,40,41], justifying the use of a certain percentage of steel rods.

FRC has been combined with FRP pultruded profiles or FRP systems made by a vacuum-assisted resin transfer moulding (VARTM) process for the development of very-lightweight construction systems, simpler to transport, install and dismantle, immune to corrosion and presenting low maintenance costs [8,9,10,11,12,13].

In the present work, FRC is combined with glass fibre reinforced polymer (GFRP) gratings on the production of manhole covers of different load-carrying capacities for forming a portfolio of solutions according to the classes imposed by BS EN 124:1994 [42], economically competitive with the existing solutions in the market, and without sufficient economic value that justifies being stolen. The use of GFRP gratings with concrete has already been explored in the development of new types of slab systems with interesting results [43,44,45,46,47,48,49,50], so the aim is now to take advantage of FRC to avoid the brittle behaviour of concrete when submitted to load conditions that introduce cracking risk or punching failure.

In accordance with BS EN 124:1994, the manhole covers are classified into six groups according to the load-carrying capacity demands required, which are dependent on the place where they will be installed, as indicated in

Table 1. Classification of manhole covers for urban roads and highways (BS EN 124:1994 [42]).

Class	${\mathit{F}}_{\mathit{C}\mathit{l}}$	Application
A15	15 kN	Areas where only pedestrians have access.
B125	125 kN	Car parks and pedestrian areas with only occasional vehicular access.
C250	250 kN	Car parks, industrial sites and areas with slow moving traffic, also in some highway locations.
D400	400 kN	Areas where cars and lorries have access, including carriageways and hard shoulders.
E600	600 kN	Areas where high wheel loads are imposed such as docs.
F900	900 kN	Areas where particularly high wheel load are imposed such as aircraft pavements.

and Figure 2. The first four classes of manhole covers are related to the urban roads and highways, which starts from the first class of manhole covers, A15, that are used in areas where only pedestrians have access, through to the fourth class of these elements, D400, which are applicable where cars and lorries have access. The last two classes of manholes are applied in zones with traffic of very high wheel loads, and not covered in the present work. The classes of manholes are dependent on their maximum load-carrying capacity ($F_{Cl}$), also indicated in Table 1.

To assess the potential of this new concept of manhole cover, an experimental program was carried out, and the relevant results are presented and discussed in this work.

2. Experimental Program

2.1. Materials and Structural Concept of the New Manhole Cover

For the development of the new manhole, two types of FRC materials and two types of GFRP grating were used (Figure 3). Regarding the FRC, a group of manholes were built with a mortar reinforced with two polyacrylonitrile (PAN) fibres of different lengths, and another group with concrete reinforced with steel and polymer macro fibres. Hereafter the first and the second class of FRC will be nominated by the acronyms PFRM and HFRC, respectively. In terms of GFRP gratings, two types were adopted, with differences in terms of the element’s cross-sectional area and centreline open mesh of the grating, since these variables are expected to impact the stiffness and load-carrying capacity of the manhole. One type of GFRP grating is formed by elements of 7 × 35 mm² cross section and 38 × 38 mm of centreline open mesh, hereafter designated by G_type1 (Figure 3a,b). The other GFRP grating is formed by elements of 6.3 × 50 mm² cross section and 50.8 × 50.8 mm of centreline open mesh, hereafter designated by G_type2 (Figure 3c). The G_type1 was used to produce manholes with PFRM and with HFRC, while the G_type2 was exclusively used to manufacture manholes with HFRC (Figure 3).

2.2. Test Groups

Table 2. Details of the specimens included in the experimental program (dimensions in mm).

Specimen Designation	Type of GFRP	Type of FRC	Geometry and GFRP Position (Cross-Section View)
PFRM80_c15_g35	G_type1	PFRM
PFRM100_c15_g35	G_type1	PFRM
HFRC80_c25_g35	G_type1	HFRC
HFRC100_c30_g35	G_type1	HFRC
HFRC100_c25_g50	G_type2	HFRC
HFRC100_c00_g50	G_type2	HFRC
HFRC110_c40_g50	G_type2	HFRC

presents the manhole specimens of the experimental program. This was formed by seven specimens, whose designation is A#1_c#2_g#3, where A can be PFRM or HFRC, #1 is the depth of the specimen (80-, 100- or 110 mm), #2 is the bottom cover thickness of the GFRP grating (0-, 15-, 25-, 30- or 40 mm), and #3 is the depth of the GFRP grating (35- or 50 mm). For instance, HFRC100_c30_g35 represents a manhole made by HFRC and G_type1 GFRP grating (g35) of 100 mm of depth and a bottom cover thickness for the GFRP grating of 30 mm (c30).

Testing the PFRM80_c15_g35 and PFRM100_c15_g35 is intended to assess the influence of the thickness of the manhole on its structural response when using PFRM and G_type1 GFRP grating. The influence of the GFRP type when using HFRC to produce the manholes is intended to be derived by testing HFRC100_c30_g35 and HFRC100_c25_g50. The influence of the FRC material and concrete cover thickness of the GFRP is expected to be extracted by testing the following two groups: (1) PFRM80_c15_g35 and HFRC80_c25_g35; (2) PFRM100_c15_g35 and HFRC100_c30_g35. The influence of the thickness of the manhole specimens when using HFRC is assessed by testing HFRC80_c25_g35 and HFRC100_c30_g35. Finally, the influence of the concrete cover thickness when using G_type2 and HFRC is assessed from the results of HFRC100_c25_g50 and HFRC100_c00_g50.

Preference was given to test seven different solutions of almost real-scale manhole cover prototypes (despite the possibility of some dispersion occurring in the experimental results) instead of testing a higher number of smaller specimens of two or more twins and then dealing with the ambiguity of transforming these results to the ones expected to be obtained in real dimension prototypes for considering the size effect.

2.3. Test Setup and Monitoring System

Since the supplier of the GFRP gratings did not provide their material properties, two representative samples of G_type1 and G_type2 were tested according to the setup shown in Figure 4a. The modulus of elasticity of the GFRP gratings was obtained from the force vs. centre deflection in these experimental tests, by simulating numerically these tests, a subject covered in Section 2.4.2. This test setup includes a supporting system for the specimens with a geometry according to the recommendations of BS EN 124:1994 [42], Figure 5. This supporting system was made by HFRC with a clear opening of 600 mm (Figure 5a), containing four additional openings near its base for providing access to its interior for the installation displacement transducers (LVDTs) to measure the deflection of the specimen (Figure 5b). The support is complemented with a 5 mm thick steel sheet that covers its top surface (Figure 5a,b) to avoid direct contact between the specimen and the supporting concrete block, preventing any possible deterioration in the referred surface due to abrasion/friction effect during the tests.

Each test was carried out using a servocontrolled testing machine, with a load-carrying capacity of 500 kN. The internal LVDT of the actuator was used to control the vertical displacements during the test. The test procedure consisted of applying a monotonic load in a circular steel plate of 250 mm diameter and 100 mm thickness, disposed in the centre of the specimen (Figure 4). To also record the response of the specimens in their post-peak load-carrying phase, the loading was displacement controlled at a rate of 20 μm/s up to their failure.

Figure 4b and Figure 6 show the test setup adopted for testing the manholes indicated in Table 2. It is practically the same adopted for testing the GFRP gratings, with the exception that the deflection of the manhole specimens was recorded by two external LVDTs disposed of according to the schematic representation of Figure 4a,b: LVDT1 in the centre of the specimen and LVDT2 at 170 mm distance from LVDT1.

2.4. Properties of Constituent Materials and Reinforcing Systems

2.4.1. Fibre-Reinforced Cementitious Materials

Two types of FRC were used in the production of the manholes: cement-based mortar reinforced with PAN fibres (PFRM) and hybrid fibre-reinforced concrete (HFRC). To produce the PFRM, PAN fibres of 6 and 12 mm length, PAN6 and PAN12, respectively, were used with the properties indicated in

Table 3. Properties of the fibres used in the production of the PFRM and HFRC.

Type of Concrete	PFRM		HFRC
Fibre type	Polyacrylonitrile fibres		Hooked end steel fibres	Polymer macro fibres
	PAN6	PAN12
Length (mm)	6	12	33	54
Diameter (mm)	0.058	0.026	0.50	Not available
Elasticity modulus (MPa)	9910	6856	210,000	7000
Tensile strength (MPa)	564	264.4	1100	450
Density (g/cm3)	1.17	1.17	7.9	0.91
Elongation	13–17	14–18	-	-

.

The HFRC was reinforced with 90 kg/m³ hooked end steel fibres and 3 kg/m³ of polymer macro fibres, whose relevant properties are indicated in Table 3.

The production technology of the PFRM is described elsewhere [8]. The mix composition of PFRM is indicated in

Table 4. PFRM and HFRC mix compositions.

Composition	PFRM	HFRC
Cement (kg/m3)	535	462
Fly Ash (kg/m3)	535	139
Limestone Filler (kg/m3)	102	139
Water (L/m3)	415	197
Superplasticizer (L/m3)	32	15.7
Fine river sand (kg/m3)	214	126
Coarse river sand (kg/m3)	-	670
Coarse aggregate (kg/m3)	-	512
Steel fibres (kg/m3)	-	90
Synthetic macro fibres (kg/m3)	-	3
PAN fibres (kg/m3)	46.8	-
Viscosity Modification Agent (L/m3)	0.825	-

, where it is visible that the aggregates are limited to fine river sand, and a content of fly ash equal to the cement content was adopted. As a result of an optimization process for the PAN-type fibres to be adopted in the PFRM, a fibre volume of 4% was obtained, composed of 3% of PAN12 and 1% of PAN6. This material was selected for producing a group of manholes (PFRM80_c15_g35 and PFRM100_c15_g35) due to its self-compacting nature, fibres and aggregates were much smaller than the centreline open grid of the G-Type1, therefore the probability of segregation of these material constituents occurring when flowing through this GFRP grating was very small. This PFRM presented a deflection hardening behaviour in tests with un-notched specimens under bending [6], which was assumed to be very convenient for the stiffness of the manholes in their post-cracking phase due to the favourable tension–stiffening effect provided by PFRM [51]. In fact, the flexural strength was much higher than the flexural stress at crack initiation, as a consequence of the fibre-resisting mechanisms that promoted the formation of a diffuse crack pattern [8]. This behaviour has the potential to better distribute the tensile forces along the elements of the GFRP grating, postponing rupture localization for higher load levels on the manhole.

The production technology of the HFRC is described elsewhere [52], and its mix composition is included in Table 4. This HFRC presented self-compacting nature, with a spread diameter of 500 mm within 3.5 s, and a total spread diameter of 660 mm [53]. No visual sign of segregation was detected in the developed HFRC, and the mixture presented good homogeneity and cohesion while flowing through the Abrams cone.

By executing compressive tests according to the ASTM C 109/C109M-11b [54] recommendations with four cubic PFRM specimens of 50 mm edge of 28 days of age, an average compressive strength of 24 MPa was obtained. An average Young’s modulus of 10 Gpa was registered by performing tests according to the ASTM C469 [55] recommendations with four cylindrical specimens of 50 mm diameter and 100 mm height at 28 days of age. To characterize the flexural behaviour of the developed PFRM, four-point bending tests were executed with specimens of 250 × 80 × 18 mm³ (length × width × thickness). The span length of these specimens was 230 mm, with 80 mm between the reaction support and its closest applied load. More details on the test setup, monitoring system, and loading conditions are presented elsewhere [8]. In these tests, the developed PFRM presented a pronounced deflection hardening behaviour and high energy absorption capacity, with an average flexural strength of 7.85 MPa at a peak deflection of 3.48 mm.

The mechanical performance of the HFRC at hardened state was assessed by evaluating Young’s modulus (BS EN 12390-13 [56]), the compressive strength (ASTM C39/C39M-14a [57]), and the flexural behaviour (Model Code 2010, MC2010 [58]) at 28 days of age. The average compressive strength, $f_{cm}$, and Young’s modulus, $E_{cm}$, of three HFRC cylindrical specimens of 150 mm diameter and 300 mm height were 57.3 MPa and 35,583 MPa, respectively. The flexural behaviour of the HFRC was assessed by determining the limit of proportionality, $f_{fct,L}$, which is the flexural stress up to a crack mouth opening displacement (CMOD) of 0.05 mm, and the residual flexural tensile strength parameters ($f_{Ri}$, with j = 1 to 4). The $f_{R1}$, $f_{R2}$, $f_{R3},$ and $f_{R4}$ correspond to CMOD of 0.5-, 1.5-, 2.5-, and 3.5 mm, respectively. For this purpose, three point-bending tests on three simply supported notched beams of 150 × 150 × 600 mm³ were carried out according to the recommendations of MC2010. The results are represented in

Table 5. Limit of proportionality and residual flexural tensile strength of the developed HFRC.

	${\mathit{f}}_{\mathit{f}\mathit{c}\mathit{t},\mathit{L}}$ [MPa]	${\mathit{f}}_{\mathit{R}1}$ [MPa]	${\mathit{f}}_{\mathit{R}2}$ [MPa]	${\mathit{f}}_{\mathit{R}3}$ [MPa]	${\mathit{f}}_{\mathit{R}4}$ [MPa]
Average	13.63	13.23	13.09	11.73	10.22
CoV (%)	23.6	12.45	10.76	11.06	15.89

, where it is verified that the residual flexural tensile stresses at a CMOD of 0.5 mm ($f_{R1}$) and 2.5 mm ($f_{R3}$) were 13.23 MPa and 11.73 MPa, respectively. Considering the coefficient of variation for these values, the number of tested specimens, and the recommendations of MC2010, the characteristic values of $f_{R1}$ and $f_{R4}$ are 10.12 MPa and 9.28 MPa, respectively; therefore, this HFRC is classified as “10c” for the toughness class according to this guideline.

2.4.2. GFRP Gratings

By testing two exemplars of the two types of GFRP gratings according to the setup described in Section 2.3, the load vs. central deflection relationship depicted in Figure 7 was obtained. The results evidenced that the GFRP grating with 50 mm height, G_type2, was as expected, presented a higher stiffness and load-carrying capacity than the GFRP grating with 35 mm height, G_type1, but of smaller deflection at failure. The average maximum load and corresponding deflection of G_type1 was 79 kN and 28 mm, while, in G_type2, the values of 99 kN and 21 mm were obtained.

Figure 8 shows the typical failure mode registered in these tests. It is verified that the fundamental failure mode is the in-plane delamination due to the inexistence of fibres in the vertical direction of the components forming the grating. It should be noted that the elements of this grid-type structure are subjected to shear forces, bending moments, and some torsional moments. Despite the apparent double symmetry of the testing specimens, geometric imperfections of the tested GFRP gratings introduce some torsional moments, which, combined with shear forces and bending moments, promote the observed failure mode.

The Young’s modulus of the GFRP material constituting these G_type1 and G_type2 was evaluated by inverse analysis, simulating numerically these tests with the version 4.0 of FEMIX computer program [59]. For this purpose, the G_type1 and G_type2 specimens were modelled using 112 plane shell finite elements of eight nodes and 2 by 2 integration points for the evaluation of the membrane, bending, and shear stiffness components, and assuming linear–elastic behaviour for the GFRP (Figure 9). The formulation is based on the Reissner–Mindlin theory, whose implementation in FEMIX is described elsewhere [60]. Loading and support conditions for the specimens were modelled in accordance with the experimental test setup (Figure 4a). By fitting as best as possible the average force–deflection relationship registered in the GFRP gratings with this inverse analysis, a modulus of elasticity of 15 GPa and 11 Gpa was obtained for the G_type1 and G_type2, respectively. The decrease in the modulus of elasticity of the GFRP with the increase in the depth of the cross-section of the GFRP grating might be a consequence of the larger number of layers required to produce G_type2 over G_type1. Since this layered organization was not simulated numerically, the localization of damage in these interfaces can justify the obtained values; therefore, they should face a modulus of elasticity of the GFRP in the context of assuming the GFRP grating as a linear and isotropic material. More complex simulations could be performed, even considering the elements of the GFRP grating as an assemblage of unidirectional GFRP layers, but the lack of information to simulate the bond conditions between two consecutive layers would mean that this exercise is not useful.

3. Results

3.1. Relevant Results, Failure Modes and Classes

Table 6. Relevant results from the experimental program.

Specimen Designation	${\mathit{P}}_{\mathbf{max}}$ [kN]	${\mathit{u}}_{{\mathit{P}}_{\mathbf{max}}}$ [mm]	${\mathit{K}}_{\mathit{T}\mathit{i}}$ [kN/mm]	${\mathit{u}}_{\mathit{p}\mathit{d}}^{\mathbf{exp}}$ [mm]	${\mathit{u}}_{\mathit{p}\mathit{d}}^{\mathit{a}\mathit{d}\mathit{m}}$ [mm]	Class According to [42]
PFRM80_c15_g35	122.97	9.83	17.22	0.0	6.0	A15
PFRM100_c15_g35	157.65	4.29	37.22	0.0	6.0	B125
HFRC80_c25_g35	208.10	32.71	33.13	0.8	6.0	B125
HFRC100_c30_g35	349.96	8.32	49.12	0.0	2.0	C250
HFRC100_c25_g50	311.65	17.36	31.66	0.3	2.0	C250
HFRC100_c00_g50	478.08	7.62	68.88	0.0	2.0	D400
HFRC110_c40_g50	423.87	15.40	46.83	1.2	2.0	D400

presents the relevant results obtained on the seven tested manhole specimens, namely, maximum load ($P_{max}$), centre deflection at this maximum load ($u_{P_{max}}$), and initial tangent stiffness ($K_{Ti}$). According to BS EN 124: 1994, the class of the manhole cover depends not only on its maximum load-carrying capacity ($P_{max}$) but also on the experimentally registered permanent deflection ($u_{pd}^{exp}$) over the admissible permanent deflection ($u_{pd}^{adm}$). The $u_{pd}^{adm}$ is equal to $1/L·CO$, where $CO$ is the clear opening of the specimen’s support (600 mm, Figure 5a), while $L$ is 100 for the manholes of class A15 and B125, and $L$ = 300 for the other classes. To determine $u_{pd}^{exp}$, to the displacement at $F_{2/3Cl}=2/3·F_{Cl}$ in the force–deflection relationship registered experimentally ($u_{2/3Cl}^{exp}$) is subtracted the recuperated elastic deformation ($u_{el}^{exp}=F_{2/3Cl}/K_{Ti}$) due to the unloaded phase: $u_{pd}^{exp}=u_{2/3Cl}^{exp}-u_{el}^{exp}$. The values of $u_{pd}^{exp}$ and $u_{pd}^{adm}$ are indicated in Table 6, from which is determined the manhole class, which is included in the last column of the table. It is verified that all manhole specimens of 100 mm depth made by HFRC attained class C250 or higher class. Despite the HFRC100_c30_g35 has included G_type1 grating (smaller stiffness and load-carrying capacity), it attained Class C250, to which it has contributed the relatively large top HFRC cover thickness (35 mm). This characteristic may justify the fact that HFRC100_c00_g50 has attained class D400, since it has the highest top HFRC cover thickness (50 mm). The HFRC100_c00_g50 has null bottom HFRC cover thickness, therefore it disposes of the highest internal arm for the reinforcement composed by the G_type2 grating, which may have contributed to the highest structural performance of this specimen. Since this reinforcement is immune to corrosion, its disposition without the bottom cover is possible, which, when combined with the highest top HFRC cover thickness, had a significant impact on the manhole load-carrying capacity (the highest of the experimental program). The influence of the top HFRC cover thickness was also visible when comparing the results of HFRC100_c30_g35 and HFRC100_c25_g50. Despite the G_type1 grating has been adopted in the HFRC100_c30_g35, the top HFRC cover thickness with more than 10 mm than the one in the HFRC100_c25_g50 has contributed to the larger stiffness and load-carrying capacity registered in HFRC100_c30_g35. The class D400 was also attained by HFRC110_c40_g50, which, despite having the highest bottom concrete cover thickness (40 mm), an increase of 10 mm on the manhole’s thickness (110 mm) allowed a peak load of 423.87 kN.

All tested manhole specimens failed by punching, with the typical failure shown in Figure 10. The delamination of the elements forming the GFRP grating is visible, revealing the lack of fibres in the orthogonal direction to the longitudinal axis of these elements. The typical pattern of diffuse circumferential cracks on the top surface of the manhole specimens is also visible, indicating the high fibre reinforcement effectiveness in arresting the propagation of cracks that they cross, delaying the damage localization due to punching.

Figure 11 shows the crack pattern on the bottom face of all tested manhole specimens at their failure. The formation of radial cracks is visible, typical of a flexural failure of an RC slab supported on a circular perimeter, but the propagation of these cracks was interrupted by the governing circumferential punching failure surface.

In the following sections, the influence of some variables considered in the experimental program is analysed regarding the structural performance of the tested manhole specimens.

3.2. Influence of the Thickness of the Manhole Specimens in Series with PFRM

The influence of the thickness of the manhole specimen on its structural response when using PFRM reinforced with G_type1 was assessed by testing PFRM80_c15_g35 and PFRM100_c15_g35 specimens, whose force–deflection is compared in Figure 12. The $P_{max}$ and $K_{Ti}$ of PFRM80_c15_g35 and PFRM100_c15_g35 were, respectively, 123 kN and 17 kN/mm, and 158 kN and 37 kN/mm (Table 6). It is verified that a decrease of 20 mm in the thickness of the specimen (20%) had a significant impact in terms of initial tangent stiffness with a reduction of about 54% and on the peak load, with a reduction of 22%. This structural performance has also contributed to the 20 mm more PRFM top cover thickness in the PFRM100_c15_g35, which is relevant due to the punching failure mode governing the behaviour of these manhole specimens. Nonetheless, the PFRM80_c15_g35 presented a high level of post-peak load-carrying capacity, with 94% of its peak load at a deflection of 25 mm. This ultimate load-carrying capacity was only 13% smaller than the ultimate load-carrying capacity of the PFRM100_c15_g35. Despite the much higher abrupt load decay that occurred on the PFRM100_c15_g35 just after its peak load, the load-carrying capacity at failure was still 84% of the peak load. Therefore, despite the GFRP grating used as the main reinforcement having a linear brittle behaviour (Figure 7), the manhole specimens developed a high level of pseudo-ductility (“pseudo” since a yield initiation of the reinforcement, typical of concrete structures reinforced with steel systems, is not applicable to the present case).

3.3. Influence of the GFRP Grating and Top Concrete Cover Thickness When Using HFRC

Figure 13 compares the force–deflection registered in the HFRC100_c30_g35 and HFRC100_c25_g50 in order to assess the influence of the type of GFRP grating and top concrete cover thickness. The HFRC100_c30_g35 uses the G_type1 with a 35 mm depth and has a 35 mm top cover thickness, while HFRC100_c25_g50 employs the G_type2 with 50 mm depth and 25 mm top cover thickness. Both manholes are produced with HFRC. The maximum load and initial tangent stiffness of HFRC100_c30_g35 and HFRC100_c25_g50 were, respectively, 350 kN and 49 kN/mm, and 312 kN and 32 kN/mm. Therefore, despite the use of G_type2 of larger stiffness in the HFRC100_c25_g50, the smaller top cover thickness (25 mm versus 35 mm in the HFRC100_c30_g35) has governed the stiffness and load-carrying capacity, providing an increase in terms of peak load and initial stiffness of 12% and 53%, which has favourable repercussion in terms of SLS design verifications for deflection. It is also verified that, while the HFRC100_c30_g35 has maintained an almost constant residual load-carrying capacity between 15 mm and 40 mm (≅70% of the peak load), the HFRC100_c25_g50 developed a gradual decrease in load-carrying capacity above the deflection corresponding to peak load, with almost null capacity at about 50 mm of deflection. Figure 11e evidences that, at failure, the damage in the HFRC100_c25_g50 was more intense than in the HFRC100_c30_g35 (Figure 11d), as was expected due to the almost null load-carrying capacity of this specimen at the end of testing, which is a consequence of the smaller top HFRC cover thickness in the HFRC100_c25_g50 for resisting the punching loading conditions. Despite the similar crack pattern, the damage was more intense in the HFRC100_c30_g35. The relatively small difference in the bottom HFRC cover thickness between these specimens (30 mm vs. 25 mm) does not provide a significant difference in the geometric centre of the GFRP reinforcement; therefore, this aspect would have not contributed significantly to the performance of these specimens.

3.4. Influence of the FRC Material

To assess the influence of the type of FRC material, the force–deflection of PFRM80_c15_g35 vs. HFRC80_c25_g35 is compared in Figure 14a, and PFRM100_c15_g35 vs. HFRC100_c30_g35 in Figure 14b. Although the top and bottom concrete cover thickness have an impact due to the reasons already pointed out, mainly the top concrete cover thickness, the FRC type material, had a higher influence on the load-carrying capacity of these specimens. In fact, when using HFRC instead of PFRM, an increase of 69% in the $P_{max}$ was registered in the specimens of 80 mm thickness (123 kN in PFRM80_c15_g35 and 208 kN in HFRC80_c25_g35), while an increase of 121% in the peak load was observed in the specimens of 100 mm thickness (158 kN in PFRM100_c15_g35 and 350 kN in HFRC100_c30_g35). The type of FRC has also influenced the $K_{Ti}$. In fact, when using HFRC instead of PFRM, an increase of 92% in the $K_{Ti}$ was registered in the specimens of 80 mm thickness (17 kN/mm in PFRM80_c15_g35 and 33 kN/mm in HFRC80_c25_g35), while an increment of 32% in the $K_{Ti}$ was observed in the specimens of 100 mm thickness (37 kN/mm in PFRM100_c15_g35 and 49 kN/mm in HFRC100_c30_g35). The smaller increase in this last comparison can be attributed to the larger top concrete cover thickness and internal arm of the GFRP grating in PFRM100_c15_g35 over HFRC100_c30_g35.

3.5. Influence of the Thickness of the Manhole Specimens in Series with HFRC

The influence of the thickness of the manhole specimens when using HFRC is assessed by comparing in Figure 15 the force–deflection of HFRC80_c25_g35 and HFRC100_c30_g35, both reinforced with the G_type1 grating. It is verified that an increase of 20 mm on the manhole specimen thickness (20%) has provided an increase in $P_{max}$ and $K_{Ti}$ of 68% and 48%, respectively. This increased level can also be attributed to the larger top HFRC cover thickness of the HFRC100_c30_g35. Above a defection corresponding to about 50 kN of load level, the HFRC80_c25_g35 experienced a significant degradation of stiffness up to the peak load, to which it has contributed the relatively small top HFRC cover thickness for facing the intense damage due to punching (Figure 11c). This manhole specimen presented a relatively small load decay after the first peak load (207.51 kN at a deflection of 15.46 mm), corresponding to the punching of the top HFRC cover thickness. This was followed by a smooth deflection hardening up to the peak load (208.1 kN at a deflection of 32.71 mm) due to the contribution of the GFRP grid and surrounding HFRC. The peak load corresponded to the rupture of the GFRP grating, followed by a pronounced load decay. The rupture of the GFRP grating of HFRC100_c30_g35 occurred at a larger load and deflection, mainly due to the larger bottom HFRC cover thickness and internal arm of the GFRP grating of this manhole specimen.

3.6. Influence of the Concrete Cover Thickness in Series with HFRC and G_type2 Grating

The influence of the HFRC cover thickness in the series of manhole specimens of equal depth (100 mm) reinforced with Gtype2 grating is assessed by comparing the force–deflection of HFRC100_c00_g50 and HFRC100_c25_g50 (see Figure 16); the first one has 50 mm and 0 mm top and bottom cover thicknesses, respectively, while the second has 25 mm thicknesses of the top and bottom covers. With a 0 mm bottom cover thickness in the HFRC100_c00_g50, the internal arm of the GFRP grating is also higher than in HFRC100_c25_g50. These aspects have provided an increase of 53% and 118% on the $P_{max}$ and $K_{Ti}$ of HFRC100_c00_g50 over HFRC100_c25_g50. Despite the much larger peak load provided by the thicker top HFRC cover due to its higher punching resisting capacity, just after peak load, the HFRC100_c00_g50 has experienced an abrupt load decay up to a load level like the one of HFRC100_c25_g50, and since then both specimens developed similar behaviour. This relatively high loss of load-carrying capacity may have contributed to the deficient infilling of some cells of the GFRP grating (Figure 11f). Therefore, for the case of 0 mm bottom cover thickness, whose advantage is the larger arm of the GFRP grating, it is recommended to optimize the HFRC’s rheology for perfect filling of the GFRP’s cells. The obtained results enforce the role of the top cover thickness on the structural performance of this type of manholes at SLS conditions.

4. Conclusions

In this work, seven manhole specimens were experimentally tested under displacement control up to their rupture. The following variables were investigated: type of fibre-reinforced cementitious material (FRC); type of GFRP grating; depth of the specimens; and top and bottom cover thickness of the GFRP grating. Two types of FRC were considered: mortar reinforced with two types of PAN fibres (PFRM); and concrete reinforced with hooked ends steel fibres and polymer macro-fibres (HFRC). The HFRC presented higher stiffness, compressive strength, and post-cracking tensile capacity than the PFRM. Two types of GFRP gratings were adopted, G_type1 and G_type2, the last one with a higher stiffness and load-carrying capacity. For the depth of the manhole specimens, 80, 100, and 110 mm were considered. Regarding the top cover thickness of the GFRP grating, values between 20 and 50 mm were considered, while, for the bottom cover thickness, values between 0 and 40 mm were adopted. From the experimental program the maximum load ($P_{max}$), the centre deflection at this maximum load ($U_{P_{max}}$), the initial tangent stiffness ($K_{Ti}$) and, based on the maximum load, the corresponding class according to BS EN 124: 1994 [42] were obtained. Based on the obtained results, the following conclusions can be pointed out:

• Regarding the influence of the thickness of the manhole specimens in series made by PFRM, an increase of 20 mm in the thickness (20%) has increased the $K_{Ti}$ by 54%, and the $P_{max}$ by 22%. The increase in the thickness was concentrated on the top PFRM cover thickness, demonstrating the influence of this characteristic on the performance of this series of manhole specimens.
• By testing manhole specimens of the same thickness but different GFRP gratings and top cover thicknesses in the series with HFRC it was verified that the top HFRC cover thickness had a higher impact on the structural performance than the stiffness and load-carrying capacity of the adopted GFRP gratings. This is a consequence of the dominant punching failure mode. It was verified that an increase from 25 mm to 35 mm on the top HFRC cover thickness has provided an increase of 12% and 53% in the $P_{max}$ and $K_{Ti}$, respectively.
• The use of FRC instead of PFRM has provided an increase of 69% and 92% in terms of $P_{max}$ and $K_{Ti}$, respectively, in the series with manhole specimens of 80 mm depth; while, in the series with manhole specimens of 100 mm depth, the increase in $P_{max}$ and $K_{Ti}$ was 121% and 32%, respectively.
• In the series made by HFRC, an increase of 20 mm in the manhole specimen thickness (20%) has provided an increase in $P_{max}$ and $K_{Ti}$ of 68% and 48%, respectively. This depth increase was exclusively due to the increase in the top HFRC cover thickness, revealing the influence of this parameter.
• By taking advantage of the immunity of GFRP to corrosion, the highest structural performance of a manhole specimen is obtained by disposing of the G-type2 with null bottom HFRC cover thickness and the largest possible top HFRC cover thickness. Even for a manhole specimen of 100 mm depth (≅24 kg/m² of deadweight), its class was D400, which means that it can be installed in the highest loaded zones of a highway.
• All manhole specimens failed by punching. Therefore, the stiffness and load-carrying capacity of the tested manhole specimens were dependent on the top concrete cover thickness of the GFRP grating and the type of FRC. In fact, these structural performance indicators have increased with the thickness of the top concrete cover thickness, and with the stiffness and resistance of the FRC.
• Since $K_{Ti}$ and $P_{max}$ increased significantly with the top concrete cover thickness, this characteristic should be taken into account regarding its favourable repercussion in terms of SLS design verifications for deflection.
• All the manhole specimens developed a high level of pseudo-ductility.

The experimental program has only considered one test per type of manhole cover proposed, which is insufficient since the level of the dispersity of the relevant results is not known. Therefore, more experimental tests are recommended to be executed to obtain results of statistical representativeness for the development of a reliable design procedure.