Large-scale Additive Manufacturing for Rapid Vehicle Prototyping

July 1, 2016

In model-based development of vehicle powertrains, through hardware-in-the-loop (HIL) to mule integration, a new enabling design tool is emerging from recent advances in large-scale additive manufacturing (AM) that has become known as big area additive manufacturing (BAAM). AM creates components directly from a computer model and is well-suited for rapid prototyping as it is extremely flexible and enables the rapid creation of very complex geometries with minimal waste. This technology could be transformative for many sectors including automotive.

Until recently, AM processes were constrained to relatively small scales for both polymers and metals. The polymer AM processes used in these applications and studies have been limited in scale due to the constraint of needing reduced oxygen and constant heat environments. In addition, there are some issues with residual stresses during the AM process with both metals and polymers that have made larger-scale printed parts difficult to produce with precision and dimensions needed for automotive applications.

Recent advances in BAAM with polymers and composites have enabled larger scales. Conventional polymer additive systems are capable of producing workpieces in the size range of less than a few cubic feet in volume. The BAAM systems from Cincinnati Inc. now have the ability to print pieces on the order of 1000 ft3 (28.3 m3). The initial BAAM systems were the result of co-development by Oak Ridge National Laboratory (ORNL) and Lockheed Martin.

The ability to directly print a large, complex working part directly from a CAD file with these BAAM systems allows for direct generation of parts such as welded tube frames and body-in-white as opposed to conventional mule manufacturing processes. To date, however, there have only been limited efforts to use large-scale AM for vehicles. For example, in 2014, ORNL, Cincinnati and Local Motors printed a vehicle named the Strati at a trade show using a BAAM system.

This article focuses on the combined use of a BAAM system and HIL for rapid vehicle prototyping. HIL is deployed to develop the powertrain and its controls with the same accelerated time frame achieved by BAAM to create the vehicle chassis. Opportunities and challenges associated with the use of BAAM for rapid prototyping of vehicles are documented, using a printed Shelby Cobra replica as the case study.

BAAM for printed vehicle prototype

The new BAAM system is able to print polymer components at speeds 500 to 1000 times faster and 10 times larger than is possible with current industrial additive machines. These systems are able to produce very large components, including those on the order of a vehicle frame, from pellets and provide a unique resource for rapid vehicle prototyping.

The BAAM system used for this project was a polymer-extruded nozzle fitted to a multi-axis computer-aided servo system. Other key features include the use of a 0.2-in (5-mm) diameter nozzle, resulting in a 0.03-in (0.76-mm) surface variation. The BAAM system was capable of deposition rates of about 20 lb/h. In addition, the pellets used for printing were relatively inexpensive, typically under $5/lb.

A carbon-fiber-reinforced ABS plastic was used. Previous experiments had determined that a blend of carbon fiber higher than 15-20% led to significant reduction in warping out of the oven. This behavior makes the addition of carbon fiber an enabling technology for large printed workpieces and can eliminate the need for additional ovens to prevent curling.

The mechanical design of the vehicle body and frame was driven by the drivetrain, suspension and battery components already selected and by the special requirements of 3D printing with the carbon-fiber-reinforced ABS polymer. A number of different body styles were considered before settling on the classic Shelby Cobra roadster shape. Dassault Systèmes SolidWorks design software was used for all mechanical modeling.

The vehicle frame was designed specifically for 3D printing that keeps stresses below 1000 psi (6.9 MPa) and provides fasteners to prevent possible delamination between the horizontally printed layers. FEA shows that with the frame loaded to 2000 lbf (1000 lbf per side), near mid-span the stress is under 600 psi (4.1 MPa), and maximum deflection is under 0.35 in (8.9 mm). The polymer frame mass is about 260 lb (118 kg).

The polymer frame could not be designed with sufficient torsional stiffness within the envelope restrictions, so a metal torsion bar was added between the firewall and rear of the cockpit. The vehicle body consists of front, middle, and rear deck single bead (0.22-in thick) sections with multiple bonded stiffeners and internal supports. The front and rear sections are removable for maintenance access. All of these body panels were printed using the BAAM system and have a combined mass of about 430 lb (195 kg).

CAD files were transformed into STL files and input into a slicing program that transformed the 3D geometry to machine tool path commands. Printing the frame was completed in one process taking approximately 12 hours. This rate is 500 times above that normally associated with polymer AM processes. It should be noted that a newer BAAM co-developed by ORNL and Cincinnati is capable of printing larger pieces at a maximum deposition rate of 100 lb/h.

The skins took about 8 hours, and the supports 4 hours to print.

As with any layered process, BAAM exhibits anisotropic mechanical properties. The carbon fiber aligns with the tool path direction, providing manufacturing-controlled strength and stiffness. The weakest direction of the parts was between layers. To help with the integrity of the frame, drivetrain components were attached with threaded rods that put the layers in compression.

HIL development and integration of hardware

While the entire frame and body of this vehicle were printed, the powertrain, suspension and components were conventional. Vehicle systems simulations were used for component sizing for the Cobra’s electric powertrain. Electrical consumption over multiple drive cycles was used as the baseline to determine the required energy storage system (ESS) capacity. Due to the accelerated nature of the project, components under consideration were limited to commercially available powertrain components using CAN communications.

The capabilities at ORNL’s National Transportation Research Center, where vehicle modeling and testing were performed, include powertrain test cells with two 500-kW transient dynamometers suitable for Class 8 truck powertrain testing and a component test cell designed to handle smaller, individual components such as engines or traction motors. They share a 400-kW ESS and each is equipped with a dSPACE HIL real-time platform.

Following the simulation study and component selection, as components became available they were installed in the component test cell. At the same time, yet-to-be-procured vehicle components and the vehicle chassis were modeled on the HIL real-time platform.

The next phase was powertrain-in-the-loop testing, where all electric drive components were physically installed in the dynamometer cell. This included the motor, inverter, battery pack, dc-dc converter, high voltage distribution box, vehicle supervisory controller, driver interface and onboard charger. The real-time platform emulated the rest of the vehicle (transmission, driveline, wheels, chassis, driver and drive cycle) and controlled the dynamometer to subject the motor and electric drive to real-world speed and loads based on the vehicle model. The powertrain-in-the-loop provided a safe and controlled environment to design, construct, debug and validate the system.

The powertrain system was then moved out of the test cell and installed in the actual vehicle, allowing for immediate operation — i.e., the vehicle was fully operational at completion of the wiring.

The front suspension was modified from a commercially available aftermarket suspension kit. The rear suspension was modified from a rearwheel-drive passenger vehicle. Modifications to the rear suspension included adding shock mounting points and a cradle for the traction motor and gearbox, which were integrated into space originally occupied by the rear differential. All structural modifications to both the front and rear suspension were done through welding of carbon steel. An aftermarket brake system was used.

Mounting was made directly to printed parts. The frame was cross-drilled with conventional drill bits for bolt-on pieces. Care was taken to avoid melting polymer during the cross-drill process. Fasteners were used for all of the polymer-to-metal interfaces.

In addition to mechanical attachments, bonding was used extensively on the printed Cobra. For polymer-to-polymer bonding, Valvoline Pliogrip was used. Some examples include bonding the nose piece of the printed Cobra to the hood, the hood to the fender well section, and the tailpiece to the rear deck lid section. The door skins were bonded to the A-pillar sections, as were a number of smaller support pieces. Bonding was also used for repair of partial delamination and for filling voids.

Additional surface finishing through machining, sanding, filling and polishing allowed for the body of the printed vehicle to be painted with an automotive-grade paint to near Class A finish. Tru-Design helped with this process. Finishing the vehicle consisted of sanding to remove loose fibers followed by a surface prep and coating to fill the ridges created by the printing process. The carbon-fiber-reinforced ABS plastic is a new material, requiring investigation of finishing and painting, and the effects of temperature swings and long-term use are still unknown.

The use of the HIL platform meant that for this case study the powertrain controls were completed in a parallel process to the printing and integration of the vehicle hardware.

Challenges and opportunities

The ability to go from CAD directly to part with minimal or no additional machining operations holds great promise for BAAM to be used in vehicle prototyping. This case study has demonstrated that as an extension of HIL powertrain development, BAAM can accelerate design to integration for prototyping vehicles. The combination of BAAM plus the HIL development cycle demonstrates the ability to use parallel processes in vehicle prototyping as opposed to more serial processes.

The ability to use BAAM in the HIL development cycle has not been fully explored here, but the potential to integrate design iterations with printing revised components shows promise. A follow-up effort to this project has been completed in which an extended-range hybrid powertrain was developed on the framework described here. Current research is focused on integrating an advanced heat engine with additively manufactured parts into the printed car to study the generation of power for vehicles and buildings.

Industry has been involved in the AM development process, and recently Local Motors indicated an interest in using the BAAM process for low-volume production of neighborhood electric to highway vehicles in local microfactories. However, the ability to directly produce complex parts through AM processes has not been fully exploited in the vehicle space.

Carbon-fiber-reinforced ABS plastic would not be considered an engineering material for direct use in a printed vehicle for the commercial market. The material is not nearly stiff enough to be used alone in the creation of a vehicle frame. It does not take point loads well and cannot be used like steel; hence, the torsional bar system for added support. Researchers at ORNL and elsewhere are investigating other polymer formulations that would be suitable as engineering materials for large-scale printed components, including vehicles.

No documented delamination occurred after final assembly, and no significant distortion in the frame or body has been observed in more than six months of operation. This is notable considering the chassis dynamometer testing and significant on-road driving time.

This study did not focus on meeting specific vehicle performance targets or address crashworthiness, vehicle lightweighting or any other consumer acceptability issues. Studies investigating the crush performance of carbon-fiber-reinforced ABS plastic are on-going at ORNL.

This article was adapted from: Curran, S., Chambon, P., Lind, R., Love, L. et al., “Big Area Additive Manufacturing and Hardware-in-the-Loop for Rapid Vehicle Powertrain Prototyping: A Case Study on the Development of a 3-D-Printed Shelby Cobra,” SAE Technical Paper 2016-01-0328, 2016, doi:10.4271/2016-01-0328.

---