Lotus Elise: Locked, Light, Glued Tight

Automotive Industries
January 1996

At first glance the Lotus Elise is almost stunningly simple. Yet itís really quite complex. The chassis went from nothing to production in under 27 months. Designed for a lightweight sports car, it will be adapted to a mid-engined supercar, and may be licensed to other automakers. Except for the door beams, there are no welds, only structural bonds and a few fasteners in peel-prone areas. Design-protected for all present and proposed safety standards, it uses 27 different extrusions, and weights just 65 kg. (143 lbs) bare.

ìThis is the start of a new concept for the motor industry,î says Dave Minter, principal development engineer on Project M111, the Lotus Elise.

ìEverything the industry has done over the past 10 years has added weight, and created a vicious cycle.î He and others on the development team feel the Elise will change that.

Up front, a composite energy absorber supports the aluminum radiator, and acts as both the air intake and front brake scoops (See photo). "It can be tuned by changing the angle of the parts, which involves nothing more than a change in a composite tool," says Richard Rackharm, senior design engineer-vehicle engineering. When barrier tested at 30 mph, the unit performed as planned. ìThere was only superficial damage to the aluminum front structure,î says Rackham. ìThe car could have been driven away.î

ìThe radiator structure is dispensable,î Says Richard Lenthall, senior analyst-vehicles at Lotus. ìIt controls the crash pulse. We probably have the most control here because things like the taper and length of its internal beams can be built directly into the mold and controlled very carefully.î The chassis extrusions offer less control over wall thickness, but are much longer.

Hydro Aluminum Automotive Structures will deliver completed frames to the Lotus assembly line in Hethel, England. The Denmark-based company worked with Lotus to design, develop, and test the chassis. (Including Lotus design chief Julian Thompson, the core team did not exceed 20 people.) ìWe ran hundreds of samples with simple joints to test the glues,î says Rackham. ìEach piece was anodized before we applied the glue and put the fastener in. Then, after it cured (40 minutes at 200 degrees C, we broke the joint to see how much it could take. That gave us a baseline on all the glues.î

Other joints were put through a salt spray regimen that simulated a 25-year life, then fatigue tested, as were two complete structures.

Less than 12 months after program inception, a rolling chassis was completed and driven on Christmas Eve, 1994. It later underwent 1,000 miles of pave testing, the equivalent, says Minter, of 100,000 road miles. Afterward it was stripped and tested for signs if fatigue or cracking. None were found.

Ciba Polymers supplies the adhesive - it looks like blue gel toothpaste when cured - which extends just past the joint as a visual bond integrity check.

ìThey are very long,î says Lenthall, ìwhich passes the loads along a larger area. Bond thickness is controlled by ridges along the seams. Every mating surface has these ridges.î Also, each extruded member (made of architectural grade 6063 aluminum) is tied to its neighbor by a tongue-in-groove joint. And Torx-head fasteners keep critical joints from peeling. Itís not uncommon for one extrusion to perform more than one function or tie into more than one of its neighbors (See photo). According to Rackham, the joint design helps guarantee the suspension pick-up points are well within the 0.5 mm tolerance allowed.

Front and rear torque boxes are connected by the long side beams. The rear unit separates the passenger and engine compartments, and contains a centrally located, 10.6 gallon fuel tank. This protects against increases in rear crash standards, and reduces the furl loadís effect on handling. The front torque cell carries the suspension pick-ups, provides space for the battery and reservoirs, and ties into the instrument panel beam. The latter is designed for either right- or left-hand drive, acts as the knee bolster, and is an interior trim surface. ìIf you try to design with extrusions in a conventional manner,î says Rackham, ìyou donít save much weight, and the structure isnít very rigid or dimensionally stable. And the cost of an extrusion die is the same no matte how many features you put on it, so you can put the details in for free.î

But how free is free?

ìThis chassis is actually a bit less than we would pay for a typical Lotus (pressed and welded galvanized steel) back-bone chassis,î says Rackham.

ìEvery time you make a weld or a bend something, labor is involved. The bend in each main beam costs about $10, and adds structural integrity. Otherwise itís a case of cutting a piece to size, putting I ton a CNC machine, and sticking it together.î

Still, extrusions arenít used in the rear subframe because they would need steel reinforcement, not meet hard point requirements, and - with a heat conductivity four times that of steel - might not stay glued together.

Therefore, the subframe is predominately spot-welded, then hot-dipped galvanized for corrosion protection, which also bonds the seams together. The holes along the top edge reduce mass, and provide an area for the material to grow into during galvanizing.

Torsional rigidity is a tuning 10,133 Nm/degree with the optional downforce producing undertray - 9,500 without. Bending stiffness is about 9.200 N/MM. This is complemented by a suspension with 60 mm of droop travel and 100 mm of bump travel, though not all of the latter is used. For racing, the car can be lowered by 50 mm, which gives a ground clearance of 3 in. with two aboard, and leaves 50 mm of bump travel. Anti-roll bars are not used, and an adjustable rear toe link adjusts compliance steer control.

This is all fine and good, but of little use in the high-volume real world, right?

Not so, say Rackham, Minter, and Lenthall. Though approximately 700 Elises will be made each year (the first yearís production is sold out), other automakers are interested in using the chassis as a basis for their own sports cars. ìI donít see why this isnít a mass production concept,î says Minter. ìTodayís production cars are m the result of yeas of process development. Taking this design to mass production is a matter of developing the process.î

To which Rackham adds: ìYou can ramp up the process simply by increasing the number of bonding fixtures. And because we use so few extrusions, you can run the plan just a little longer to create the material to build 1,000 more cars."

Automotive Industries
January 1996