Design using composite materials is a complicated process; matching up the requirements of a design brief with the offerings of materials and the limitations of a compatible manufacturing process is like threading a moving needle. Material combinations are endless, research is wide-ranging, rapid and continuous and manufacturing options are numerous and ever-changing. The flexibility of composite manufacture is such that new production methods are often invented, or existing ones modified, to better suite a specific application. Before the design process could begin a substantial amount of research into composite materials, design and manufacture was necessary. This research was intended to help determine the optimal method of creating a carabiner from composites.

Initial exploration

The feasibility of the project was tested with a number of preliminary calculations. Several calculators were created, using Microsoft Excel, to expedite the exploratory calculations. The first set of calculators used simple stress analysis to estimate the ultimate tensile stress, diameter or strength rating of a carabiner (any one of these properties can be found if the other two are defined). To simplify the calculations it was assumed that the carabiner cross section was circular and that carabiners fail in pure tension (i.e. ignoring the bending contribution due to the applied load being offset from the longitudinal axis of the spine). The error in the latter assumption is described further in the Testing and FEA section later. By measuring the diameter of a carabiner (or approximating its area by other means), then inputting its strength rating - the design stress can be estimated, this is useful when considering new designs.

Further calculators were created to estimate: the bending stress in the spine when loaded transversely, the weight of a carabiner based on its straightened length, diameter and density, the number of fibre tows needed to achieve a certain strength rating and the volume fraction based on the number of fibre tows used. The assumption of circular cross section means that these calculators are only very rough estimates, useful for quick comparisons and 'ball-park' figures. Using these calculators, combined with data on material properties, enables an estimate of how much weight can be saved with various material combinations. An estimate of the maximum weight saving potential can be determined by comparing the specific strengths of a typical high strength laminate and 7075 T6 aluminium:

(Equation.2) (Equation 3)

Where m is the mass of the carabiner body, L is its straightened length, is the stress due to an applied longitudinal load F, A is the cross-section area (the carabiner body is assumed to have constant cross-section) andis the density of the carabiner material. Combining equation 2 and 3:

(Equation 4)

Equation 4 describes how the minimum mass required to make a carabiner body of sufficient strength depends only on the specific strength of the material, the load that it will have to carry and its straightened length. As mentioned previously, this assumes that the carabiner material fails in pure tension. Using CF/cyanate ester quasi-isotropic laminate as typical example: composite laminate density is 1670 kg/m3 and the tensile strengthis 607 MPa [51]. A quasi-isotropic laminate was chosen because in reality a carabiner will need to take tensile, compressive and bending loads in both the transverse and longitudinal directions - therefore using the tensile strength of unidirectional laminate would give overly optimistic predictions. From table 3; the densityand strengthfor 7075 T6 aluminium are 2810 kg/m3 and 510 MPa respectively. Based on a typical body length of 19.5 cm and an applied load of 24 kN the mass of the carabiner body will be 13 g and 26 g for composite and aluminium respectively (table 8). Based on a wiregate mass of 6 g, this comes to a maximum percentage weight saving of 40%. This is only a very rough guide to the maximum weight saving possible, realistically it is likely that a composite carabiner must be over-designed somewhat to improve impact resistance.

Table 8: Comparing the carabiner body mass for a composite and aluminium body. See preceding paragraph for material details and assumptions.


Density kg/m 3

Tensile strength MPa

Carabiner Mass g










When attempting to design a product it is important to know who will use it, how it will be used and be aware of what similar products already exist. It is also vital to have detailed knowledge of the relevant materials, manufacturing processes and design requirements. However, once this is attained to a reasonable degree, it can be very useful to come up with ideas without worrying about satisfying the material and processing limitations and design requirements [52]. There are so many restrictions and requirements that it would be easy to critique and discard most ideas - potentially missing out on useful information. Even if an idea is not feasible, it may contain elements that could dramatically improve the end product. Figures 20 to 22 show a selection of ideas and sketches created on this basis.

carabiner design brainstorming

Figure 20: Brainstorming ideas.

Top - a double axis gate that enables the gate to open in the clockwise and anticlockwise directions. The gate could be split in two; each side rotate independently. Alternatively a mechanism could be used that sets the rotation axis depending on whether the user applies an opening force on the top or bottom of the gate. Bottom - a carabiner could be cut from a panel. The panel could be made using thermoplastic prepreg compression moulding. The prepreg could be made by automatic lay-up to orient the fibre strips optimally.

carabiner design brainstorming

Figure 21: Brainstorming ideas.

Top - some existing carabiners make use of a fixed sling, where the sling, used to attach two carabiners together to make a quickdraw, cannot be removed. This is potentially useful for a composite carabiner as, in accordance with the EN standard, it circumvents the need to pass the transverse loading requirements. This means that loading directions are more predictable which makes it easier to take advantage of the excellent tensile properties of reinforcement fibres. Bottom - a sliding gate would lessen the chances of accidental gate opening.


carabiner design brainstorming

Figure 22: Brainstorming ideas. Left - various alternatives for a single part design. If the carabiner used a single part design it could be injection moulded in a single process - greatly simplifying manufacturing and reducing labour costs. Right - exploring different arm angles and spines lengths to encourage rope relocation and ease clipping.

Proposed designs

Two proposed design options are described below. These combine the most promising materials with the most effective toughening methods from literature and employ compatible manufacturing methods that are suited to the production quantity. The designs are: a CF/PEEK prepreg based design using a tennis racket style of production and an injection moulded short fibre CF/PEEK design. The injection moulded option is explored in more detail with a 3D model representation.


The manufacturing process used to create composite tennis rackets is particularly suited to carabiner production. Like tennis rackets, carabiners need high stiffness, strength and resistance to impact. In both cases, the applied load is in a very specific in direction and is repetitive. In racket manufacture a prepreg is wrapped by hand around an inflatable tube - by using hand lay-up it is possible to highly customise the orientation of the prepreg strips to give strength and stiffness in specific directions. The tube and prepreg are inserted into a closed mould. The tube is pressurised and the mould is heated until the racket has fully conformed to the mould. The use of a CF/PEEK prepreg reduces the cycle time compared to a thermosets because it is not necessary to cure the PEEK matrix - it can be removed as soon as it is cool enough to maintain its geometry outside the mould. The process will require the carabiner to be hollow which could potentially increase the cross section beyond the maximum acceptable value - this depends on what the minimum possible size is for an inflatable tube capable of applying the required pressure.

tennis racket manufacture

Figure 23: Composite tennis racket manufacture could be adapted to make carabiners. 1: Prepreg is wrapped around an inflatable tube. 2: The tube is heat and pressurised in a mould. 3: The frame is coated in polymer [53].

A tough prepreg could be prepared by using 3D woven laminates. These are more compliant in the out of plane direction compared to unidirectional tape based laminates; this means that more of the impact energy is absorbed by the response of the structure rather than through delamination [34]. In addition, the weave pattern helps to reduce the propagation of shear and delamination cracking. This would reduce the likelihood of damage if the carabiner was dropped onto a hard surface and improve its response to the impact of stopping a falling climber. In addition, PEEK is a very tough matrix and as discussed previously - using a tough matrix brings a significant improvement in the toughness of a composite structure. PEEK also passes all of the environmental resistance requirements.

The detail in the nose and gate attachment point of the carabiner may be hard to produce in the same operation as the creation of the main body as thermoplastic prepregs do not conform well to complex mould geometries. This is due to their poor tack and drape characteristics. However, it may be possible to include inserts that are joined to the carabiner during the heating process.

Although this manufacture method could be used to produce high quality carabiners, it would require a large number of separate processes; hand lay-up, manufacture of inserts, gate manufacture, moulding and final assembly of the gate and body. Hand lay-up is skilled and time consuming work that would add great expense to the parts. It is unlikely to be economically feasible to compete against current carabiners using hand lay-up of prepregs; both the labour and materials costs would be significantly higher. The rough cost of CF/PEEK laminate is £60/kg [54] compared to £8/kg [22] for aluminium alloy (table 3). It may be possible to replace the expensive CF/PEEK material with a less expensive epoxy based prepreg - however, this is not enough to mitigate the cost of extensive skilled labour.

Injection moulded

Although the tooling costs are high, injection moulding would enable the complex geometry of a carabiner body to be made in a single process. In addition, once manufacture begins there will be minimal labour costs and production capacity will be very high thanks to fast cycle times. The mechanical properties of injection moulded composites rarely compete with laminated composites because their fibres are relatively short and their alignment is hard to control. However, the strength of most high performance laminates considerably exceeds the requirements - that is to say, an injection moulded polymer does not need to match laminate strength, it must only have a superior specific strength compared to the aluminium alloy currently in use. A promising candidate material is Victrex 90HMF40 PEEK. This is a PEEK resin filled with 40% short carbon fibres and has a tensile strength of 350 MPa and a density of 1440 kg/m 3 , giving it specific strength of 243 kJ/kg comparing favourably to the 181 kJ/kg specific strength of 7075 T6 aluminium. Properties of 90HMF40 and 7075 T6 aluminium are compared in table 9.

Table 9: Comparing properties of 7075 T6 aluminium with 90HMF40 CF/PEEK [20-22, 55]. *estimate from the CES database [56].


7075 T6


Tensile modulus (GPa)



Tensile strength (MPa)



Density (kg/m 3 )



Specific tensile strength (kJ/kg)



Fracture toughness K IC , ( MPa . m 1/2 )



Notched Izod energy (kJ/m^2)



The PEEK composite has superior specific strength, but relatively poor toughness. There are no defined requirements for fracture toughness and Izod energy; testing would be necessary to determine if the composite has sufficient toughness for use as a carabiner - ideally a prototype should be tested using a drop tower.

Mechanical properties could be improved using hybridization. Adding small (0.3-1.5 µm) thermotropic liquid crystalline polymer (TLCP) fibres to a CF/PEEK composite can significantly improve tensile strength and modulus, flexural strength and modulus and marginally improve toughness [41]. The TLCP fibres also reduce the melt viscosity; this reduces the clamping force required of the injection machine - which can reduce costs. It is also worth noting that it is possible to improve the toughness and stiffness of an epoxy resin by adding alumina particles [38]. This may not work for a PEEK matrix, but is worth further exploration. Further optimisation of toughness can be attained by careful control of the cooling rate after moulding. Rapid cooling increases matrix toughness but reduces tensile strength due to reduced crystallinity - a compromise must be met to achieve sufficient strength whilst optimising toughness.

3D modelling and rapid prototyping

Two 3D concept models were created in SolidWorks. One model is based on the 'gateless' concept (where the gate and the body are a single integrated component) mentioned in the Brainstorming section and the other is a more conventional gated carabiner - both could be manufactured by injection moulding. The gateless model was created to provide a 3D visualisation of the concept (figure 24, 25). However, it was determined that the force required to open an integrated gate sufficiently to allow a rope to pass through would cause stresses very close to flexural strength of the material. Furthermore, carabiner strength is dramatically reduced in the open gate configuration. Normally the gate would restrict the arms of the carabiner from bending - reducing stresses due to bending in the spine. Additionally, a small portion of the tensile load is normally transmitted through the gate. In the gateless design the arms of the carabiner are free to bend - resulting in greater stresses in the spine. This effect is demonstrated by the fact that most carabiners are capable of holding a longitudinal load of 24 kN with the gate latched closed, whereas when the gate is open the failure load decreases to around 8kN. Nonetheless, the gateless concept might be an interesting idea to pursue for further development; a major challenge will be achieving the transverse loading requirements.

3d model of carabiner design

Figure 24: A CAD model of the gateless concept, where the gate is integrated into the body of the carabiner - enabling gate and body to be manufactured in a single operation.

rapid prototyped carabiner

Figure 25: A rapid prototyped model of the gateless concept. The CAD model in figure 24 was converted to STL format and this was used to create a rapid prototyped model by stereo-lithography.

3d model of carabiner design

Figure 26: A CAD model of the proposed carabiner design.

rapid prototyped carabiner

Figure 27: A rapid prototyped model of the proposed carabiner design.

The proposed design makes use of a curved I-beam cross-section to make the most efficient use of material whilst making sure that there are not sudden changes in cross section that could cause stress concentrations. The gate in figure 26 is added purely to demonstrate the gate position - it is a separate component and would not be injection moulded with the carabiner body. The gate would be manufactured separately from high strength steel wire. The geometry conforms to all geometrical requirements defined in the design criteria section; however, it has not been optimised for strength and stiffness. To optimise the strength and stiffness, and thus minimize weight, it is necessary to carry out finite element analysis (FEA) on the model. Some exploratory FEA was carried out on a model of an existing carabiner to check the validity of constraints and loading conditions (see FEA and testing section). The CAD model was prepared in such a way that the geometry and cross-section can be modified easily, this enables iterative improvement of the design based on feedback from FEA results. Figure 28 shows which dimensions can be modified to control the path of the carabiner body.

cad modelling carabiner

Figure 29: The CAD model was prepared so that the geometry can be easily modified in response to feedback from FEA results.

Previous: Literature

Next: FEA