Summary 

Lilleakerbyen is a district on the west side of Oslo. The Lysaker river, which it straddles, has defined its development for generations. The family run Mustad Eiendom company has occupied the 300 acre site for 150 years, initially using the river and local resources for industry, then building the infrastructure needed to support a growing community. Despite strong transport connections to the centre of Oslo, the site lacks an urban atmosphere and the amenities to attract residents and visitors. The next decade marks a turning point as the owners transform Lilleakerbyen into a vibrant mixed-use neighbourhood.  

The Lilleakerbyen masterplan understands the importance of infrastructure, civic spaces and its most valuable asset: the river. This blue green corridor both divides and connects the site: a physical barrier but also an oasis of nature within an otherwise urban environment. It is no surprise that central to the masterplan are nine bridges: some exist, most are new proposals. To develop these bridges from lines on a map into feasible structures, an international multistage competition was held over the winter of 2023-2024. Despite staggered phasing, designing them simultaneously was a chance for interaction amongst the teams and total focus by the client. 

Moxon and Cowi teamed up for the competition having collaborated on dozens of projects over the past two decades. After passing a prequalification round they were assigned three of the central bridges, competing against another team for each. Designing three bridges in close proximity is a special opportunity. In contrast to a ‘one size fits all’ solution, the team thought each should respond to the specifics of its site and function with potential for commonalities across the bridges to help them read as a family (possibly cousins rather than siblings). 

Bridge E

The northern bridges (E and F) lie at the geographical heart of Lilleakerbyen. In this area the river banks are steep and rich in vegetation. Beyond the banks there will be a high concentration of cultural activity along both sides of the river. Bridge E in particular will play an important role in connecting the primary cultural hubs. Linking these areas poses challenges that the bridge must respond to: 

• The bridge will connect two urban spaces via a natural environment.
• One end of the crossing will be significantly higher than the other.
• The bridge needs to work for onward traveling users going in all directions. 

The proposed crossing addresses each of these requirements, resulting in a rich design that is well suited to its environment.

Both bridge abutments will appear to ‘grow’ out of the adjacent plazas appearing as promontories, funnelling users towards the river then easing onward journeys at the other end. Shaping the landings allows the bridge to be a simple straight line in plan. In contrast to the heavy, grounded abutments, the slender timber bridge will appear light, floating well above the river. In elevation the gently arcing bridge will respond to the site’s topography, rising up from the west before flattening out to the east. It will appear taught with an underslung tension element strung below like the strong of a bow. The experience of crossing should be similar to a tree top walkway where users will pause to enjoy the river views in both directions.

E and F - a pair of bridges

While the proposed design could stand alone, it is stronger alongside the neighbouring Bridge F which will sit just downstream. The two bridges serve different functions: 

• Bridge E is high and light, connecting landings at different levels.
• Bridge F is low, wide and robust, extending a strong axis for pedestrians and occasional vehicles.
• One is of the ground, the other is of the air. 

Despite serving different functions, the bridges have been conceived as a pair. Both are through beam structures with similar cross sections and therefore similar user experiences. They will have human scaled parapets extending above their structural edge beams. The abutment forms and deck surfacing could be similar, allowing the bridges to meet ground in a similar manner. Elements of the landings: paving and low edge walls, could extend far beyond the bridges to literally connect them with the wider context. Despite proximity and design cohesion, riverbank foliage will limit views of the complete scheme. The best place to appreciate the relationship between the two bridges will be above the river when standing on one bridge, looking at the other. At this point the spirit of the bridges and their connection to each other and the site will be fully understood.

Structural typology

Bridge E clear spans 43m between movement joints. Footbridges of this length require careful consideration to ensure proportions are human scaled. This is often done by supporting the deck from below via an arch or props, or above via a truss, arch or cable structure. Below deck supports were ruled out due to flood levels and bank sensitivity, while overhead structures were ruled out for clashing (physically and visually) with the dense tree canopy. The remaining viable solution is a beam. Simple, honest and elegant. 

Locating beams along both sides of the deck allows them to double as edge protection. In this case the beams extend less than 800mm above the deck to ensure the experience is open and expansive. Below the deck they extend only enough to support the transverse structure and deck buildup. The resulting beam depth on its own is undersized to span the required distance. To compensate, underslung tension elements will be stretched from end to end, propping the beams at their third points. This significantly increases performance without increasing visual weight. The resulting bridge is deceptively slender with minimal materials, cost and carbon.

Structural timber

Structural timber was chosen over steel or concrete as it feels appropriate in the natural setting while reflecting the progressive spirit of the wider development. It references the industrial heritage of both Lilleakerbyen and Norway where timber has always been an important resource. Both the structural edge beams and internal timber cladding will age gracefully as the surrounding woodland banks mature. 

Glue laminated (glulam) beams are commonly used in buildings throughout Europe. They combine the benefits of quick growing softwoods with the strength and predictability of an engineered product. By laminating smaller, overlapping elements in all three directions, beams of any width, height and length can be achieved. The only limitation is the size of the workshop and the means of transport. In addition to size versatility, shaping the beams is relatively easy: the base components (lamella) are somewhat flexible before being laminated. In this case, gently curving edge beams create a dynamic appearance when viewed from the river and a more dynamic experience when walking over the bridge. 

Structural timber should support the local supply chain with native species selected and processed as close to site as possible. Transportation and erection will influence the final design; however, the beams should be no bigger than those proposed within the surrounding development or nearby infrastructure projects. Timber’s main environmental advantage is that it takes far less energy (carbon) to build with. It also captures (sequesters) carbon while growing. If its ‘next life’ is considered, it can offset carbon invested elsewhere in the project. Another advantage is timber’s relatively light weight when compared to other structural materials. Lighter weight means easier transport and installation as well as reduced foundations. For these reasons structural timber should continue to gain popularity in the built environment including the infrastructure sector.

Secondary elements

Cross beams will also be timber, spanning between metal brackets fixed to the main beams. At regular intervals this connection would be fully integral to provide u-frame stiffness and restrain lateral torsional buckling in the main beams. The timber cross beams could be composite with the concrete deck via screwed shear connectors that will increase material efficiency and stability. The robust deck will likely include precast planks with in-situ stitches at each cross beam. If possible, concrete planks could be repurposed during the demolition of a nearby parking garage. Alternatively, with appropriate formwork, the deck could be an in-situ slab. In either scenario the concrete deck will be topped with waterproofing, screed and a non-slip surfacing system, serving users while protecting the timber beams below. 

Tension elements

The underslung tension elements provide upwards thrusts to the timber beams to minimise sagging moments in the deck. With anchorages at both ends and fabricated steel props at the third points, these elements could be composite straps (e.g. CFRP - carbon fibre reinforced plastic) or high strength steel plate or cables. The dead load of the structure can be used to tension the underslung elements by temporarily anchoring the ends of the straps/cables at the abutments with any required pre-tensioning applied before lifting the deck into place. The tension element will then lengthen into its permanent position before being attached to the beams for the permanent condition.

Deck dynamics

The 43m simply supported span is potentially susceptible to pedestrian induced vibrations, but the areas under the deck between cross beams would be good locations to include tuned mass dampers to limit accelerations. The mass of the concrete deck will be considered as passive vibration dampening.

Durability and maintenance

Modular timber lining and metal capping will protect the primary beams from weather and public wear. These elements will have a shorter design life and will be removable to also inspect and maintain the beams within. Handrails with integrated LED lighting will be attached via steel T sections that double as supports for the internal lining. Like a typical timber framed building, breathable membranes and ventilation voids will ensure moisture management around the beams. The beam outer faces and bottoms will be exposed but inclined to reduce rain splashing and promote free drainage. Deck drainage will run transversely to a central shaped gulley to keep flow away from the timber edge beams where possible. It will then flow longitudinally to the abutments for discharge into controlled soakaways.

Bridge F

Bridge F is nestled deep within the river valley. The brief for this crossing requires a ‘workhorse’ solution to complete a primary north – south axis. The span is modest, but the structure must be robust to support occasional vehicles and withstand seasonal flooding. These factors and a high volume of pedestrian traffic further support a desire for solidity. This bridge design is ‘of the ground’ literally and metaphorically through both form and materiality. The primary structure is an assembly of regionally sourced stone blocks strung together to form shallow arching beams. This timeless material will create a sureness and a feeling that the bridge has been and always will be there. Combining an ancient building material with modern technology further nods to the past and future, supporting the aspirations of the wider development.

Flooding

Sitting low in the valley, this bridge anticipates flooding and floating debris that may come with it. To maximise clearance below the soffit, most of the beam depth is placed at and above deck level. At this bend in the river water flows faster along the eastern (outer) side of the channel. The resulting fluvial erosion has created a steeper eastern bank where obstructions must be avoided. This condition has informed the asymmetric bridge arrangement. A pier feels at home along the western bank, where flow is slower and the bank slopes gently. The resulting western ‘back-span’ creates structural balance while giving the river room to ‘breathe’. In this way, asymmetry links context, structure and form to create an entirely bespoke solution.

Materiality

Good quality granite exists in the area and is ideal for a structural application such as this: strong, durable, sustainable and beautiful. Natural stone which is simply quarried, cut and shaped involves less energy (carbon) than manufactured materials such as reinforced concrete. The compressive strength of granite varies but is typically much greater than 100 MPa (significantly higher than concrete). As a natural material each quarried block is typically evaluated via visual and non-destructive testing to ensure integrity. Subtle variations in tone and texture will enhance the overall bridge appearance. 

The bridge is set out along a 1.2m grid that defines cross beam spacing and the parapet module. This is subdivided into 600mm granite blocks that achieve a balance of manufacturing / transport ease with minimal wastage when quarried. Pending input from the selected quarry this could be further optimised. Granite’s durability is evidenced by ancient stone bridges that survive with little to no maintenance. Furthermore, stone can be re-purposed when a structure is no longer needed, improving its carbon credentials. Modern advances in cutting and dressing stone with computer numerical control (CNC) technology allow variation in elements with minimal additional effort. This provides the freedom to shape a structure based on its behaviour. 

Analogy to precast segmental construction

In contrast to stone’s strength in compression, it performs poorly in tension. This is why load bearing arches have always been the best way to span in stone. While embedding tensile steel reinforcement is unfeasible (as is done in concrete), metal elements have historically been used to augment masonry in architecture. With contemporary methods it’s possible to ‘thread’ high strength steel tendons through an assembly of stone blocks to prevent tension in and between the blocks under every load condition. In this manner a bridge will behave in the same manner as a post-tensioned precast segmental concrete structure. Similar segmental stone beams are being used in buildings with increasing frequency thanks to their efficiency and carbon benefits. 

Like segmental bridge construction, in this application epoxy grout will be used between blocks while stainless steel pins will ensure a full shear connection and alignment during post tensioning. High tensile steel tendons will be ‘thread’ through the blocks within a robust corrosion protection assembly. Most proprietary systems include spirally wound strands of galvanised wires that are protected by wax within a tight-fitting high-density polyethylene sleeve. These strands are bundled within a continuous duct which passes right through the stone blocks from end to end preventing water from reaching the tendons. Offsetting the tendons from the centroid of each block according to its location along the beam enables favourable bending moments to counteract externally applied loads. This migration of tendons from bottom to top and back is reflecting in the beam’s constantly changing cross section and sculptural appearance.

An integral intermediate pier along the west bank enables a continuous, haunched structure where the tensile forces (and tendons) move to the top of the cross section. Visually the pier reads as an extension of the stone blocks above (without a gap for bearings). Infilled stainless-steel crossbeams either side will create a fully integral shear connection between deck, edge beams and pier.     

At both ends of the bridge, the concentrated tension loads in the tendons will be spread within special anchorage segments to create uniform compression forces on the adjacent blocks. This function should be expressed through materiality and detailing with the tendon ends conveniently located for initial stressing, inspection and maintenance. Below the anchorage segments, elastomeric bearings will transfer vertical loads to the abutments allowing free expansion and rotation. Like the intermediate pier, crossbeams will connect the edge beams while terminating the deck system.

Secondary elements

At 1.2m centres, stainless steel plates will be sandwiched between the stone blocks, extending vertically to form handrail supports and transversely as crossbeams. Like Bridge E, concrete panels will span between crossbeams to form a composite deck via in-situ stitching. Ultra high-performance concrete (UHPC) is being pursued for its high strength and absence of steel reinforcement (ideal in a wet environment). Again, like Bridge E, the concrete structural deck will be topped with waterproofing, screed and a non-slip surfacing system. Along both edge beams, stainless steel railings will run end to end with integrated LED lighting. 

Durability and maintenance 

As a primary artery, Bridge F is wider than other pedestrian routes with capacity for emergency vehicles. The robust upstand stone edge beams are designed to withstand vehicle impact. With lighter handrails above, they will provide a sense of security, solidity and grounding for pedestrians and cyclists without feeling too enclosed. The combination of stone and stainless steel in general will ensure maximum durability with minimum maintenance. Special care will be given to joints and water management where water or dirt accumulation could cause deterioration. Minimal maintenance elastomeric movement joints at both ends will accommodate a small range of movement. As noted above, with proper detailing the tendons should require minimal maintenance. Replacement however will be possible using industry standard methods. 

Bridge G

Of all the proposed crossings, Bridge G will be the most ambitious in terms of geometric constraints. It will also be among the most used as it creates both internal and external connections to the existing and future infrastructure network. Several requirements informed the design, the most critical being: 

• Connect landings that are over 4m apart vertically.
• Promote smooth travel by cyclists and pedestrians.
• Minimise impact on the riverbanks. 

These requirements influenced the alignment which in turn influenced the structure, the materiality and the detailing. The result is an eye-catching form that should be a joy to view and to use.

Alignment

The brief for this bridge contained the ingredients of a great design: it demands innovative problem solving with a healthy balance of fixed and open parameters. The west landing location is fixed at a height well above the eastern riverside path. A perpendicular ramped connection is too steep while a diagonal ramp to the south ‘chases its tail’ as the ground drops away. An S alignment satisfies the geometric requirements while providing a pleasant route to navigate: a meandering path encourages users to look in a range of directions. Simultaneously, bends in the deck will force cyclist to stay alert and travel at reasonable speeds (which will be critical when descending west to east).

Span configuration

While an S shaped bridge will feel effortless to use, supporting it is another story. For comparison, had a diagonal single span worked, it would be over 80m long and completely out of scale in this densely vegetated river environment. Bending a clear spanning bridge into an S would be less efficient and more complicated, hence the introduction of support props. The chosen configuration respects the sensitive environment, quickly crossing the banks perpendicularly before bending to run down the middle of the river. Props along the banks lean out to stabilize and split the bridge into three nearly equal spans.   

Given the geometric nature of this crossing combined with the ambitious structural demands, potential permutations were tested parametrically. By adjusting inputs, outputs were optimized that best satisfied the objectives. A genetic algorithm helped reveal the ideal solution (at this early project stage). While this investment may seem overly technical or academic, it created the skeleton of a tool that will enable unlimited optioneering and refinement in the future. As stated above, this method of problem solving is perfectly suited to the demands of Bridge G.

Structural Typology

Having considered a range of structural options, a closed cross section (tube) in steel best handles the twisting loads imposed by the non-linear alignment. Extruding a simple trapezoid along the curving path is both stable and buildable. A closed loop of shear flow is developed within the walls of the box, with the relatively large cross-sectional area leading to fairly small shear stresses in the plates. The form also best expresses the crossing’s spirit, boldly sweeping over the river in a continuous gesture. While a more open, skeletal articulation would be interesting, it would not perform as well and might detract from the overall statement. In this way, the S dominates and the other components compliment. 

Supports

Vertical loads on the inclined props will develop both horizontal forces and bending in these members. A bottom ‘pin’ permits slight rotation, removing moment transfer into the foundation. The pin will take the form of a ‘knuckle’ joint, rotating in both axis, transmitting axial forces without mechanical parts. The prop’s slenderness allows enough flexibility so it can act in compression without buckling. Raking micro piles will efficiently transfer the prop’s horizontal forces directly into the riverbank soil. There installation could be via plant on a barge or temporary working platform.

The main span to side span ratio will create upwards forces at the ends when the centre is loaded. Torsional loads will also cause upwards forces on one side of the deck. While these forces could be restrained by a pair of bearings, integral abutments will be more efficient and easier to maintain. Stiffness at the ends will also reduce loads on the props to keep them slim. Finally, the larger foundations needed for integral abutments will be easy to install in these locations (in contrast to the harder to access prop foundations).

Deck dynamics

The ambitious form, relatively light weight and long span mean deck dynamics could be an issue for this bridge. The first 2 modes of the model have natural frequencies at approximately 2Hz, which is around the frequency where pedestrian excitation is likely to occur. If it cannot be designed out, tuned mass dampers can be added to avoid excessive accelerations. These could be incorporated within the deck box, serviced from above via access hatches. In this instance TMDs may be more economical and less carbon intensive than making the structure overly stiff. 

Durability and maintenance

The closed box limits the areas where water can pool and flow along plate surfaces. It also avoids areas for animals to settle. Despite these efforts to minimise corrosion, a typical carbon steel structure would still require infrequent repainting. If budget allows, the soffit plates could be stainless steel to eliminate maintenance over the river. This material would give the crossing an unexpected lightness as it literally reflects its surroundings. Atop the deck plate a proprietary epoxy aggregate system will protect the structure while providing a hard-wearing slip resistant surfacing. Flowing along both edges of the deck a finely crafted parapet of inward leaning stainless-steel rods will alternate between supporting a handrail and a taller cycle railing. This simple repetitive array will highlight the crossing’s dynamic qualities when passing from one bank to the other.

Conclusion

While Bridge G lacks some similarities with its relatives upstream, it expresses the same approach to context and function. All three bridges fit comfortably in their surroundings, respond to user needs and showcase the highest level of craft. They each feature a specific material (timber, stone and steel), used efficiently and ambitiously. Their variety is analogous: 

• Bridge E is the race horse that jumps the river in a single stride.
• Bridge F is the work horse supporting heavy loads in an adverse setting.
• Bridge G is the show pony that playfully traverses obstacles with grace.

All three schemes were selected by Mustad Eiendom and should be built over the next decade.