A fundamental and never-ending challenge in automobile development is to achieve sufficient resilience to withstand the rigors of on-road use without incurring a weight penalty that might degrade both performance and fuel consumption. That challenge is becoming more acute as vehicle makers seek to satisfy low carbon emission targets that are increasingly codified in law.

For Sweden-based truck manufacturer Scania, which employs approximately 35,500 employees worldwide, those objectives are achieved through the use of an array of software systems that constitute what technical manager Mikael Thellner describes as the company's computer aided engineering (CAE) “toolkit." Thellner says that in common with much of the automotive sector, the company's basic design system is the CATIA 3D CAD modeling software from French-owned Dassault Systèmes (DS).

However, for structural design—essentially producing an optimal strength-to-weight ratio in the final vehicle particularly where key load-bearing elements are concerned—Thellner highlights the roles played by two other programs. They are the TOSCA Structure and Abaqus programs, both of which are also part of the DS Simulia product portfolio now. He describes the first as an “optimizer" package and the second as a finite element analysis (FEA) “solver."

According to Thellner, this combination of software tools has been in use at Scania for a little over three years now. He says that the tools provide an upfront optimization capability that helps ensure that when detail design work to construct the final product configuration gets underway, the outline of the specific geometry must already be established.

Design Process

One organizational challenge is to convince design engineers to use the software suite. Image source: Scania The procedure is one that starts and finishes in the CATIA environment. Initially, though, that does not involve creating the intended product geometry. Instead, says Thellner, a designer working in CATIA designates a space in which a particular structure will sit; he mentions as an example a fuel tank and the bracket that will attach it to the vehicle frame. That information will then be fed into the Abaqus software which will “fill it up with elements" so that it can be processed in the TOSCA software. The result, he says, is a solution that “will tell me where I should put material and where I should take it away."

In practical terms, what comes back from the Abaqus-TOSCA combination will be an indication of the surface of the required final part geometry in the form of a STereoLithography (stl) or IGES file. This represents “a lot of dots within space," Thellner says. A designer working in CATIA then uses that data as a starting point in the construction of 3D CAD modeling information that represents an optimal solution.

This still leaves the designer with work to do. As Thellner explains, the work carried out in TOSCA and Abaqus should ensure that the final design has the required fatigue properties. It will not, however, provide guidance on issues such as ease-of-assembly. At a minimum, it means that the designer is not starting from a metaphorical blank sheet of paper.

Thellner says he is confident that, compared with previous practice, the design process is more efficient and effective. Essentially the emphasis on “iteration" in the product development process has been shifted further forward so that most of it takes place between the worlds of CAE (specialist programs such as TOSCA and Abaqus) and CAD (the 3D modeling in CATIA) rather than between CAD and physical prototyping and testing.

Tangible Results

Indeed, Scania claims some measurable achievements. For example, a bracket optimization project aimed at reducing the amount of cast iron required by the initial design succeeded in doing so by 30%. The methodology involved "freezing" functional areas and joint spaces to connecting areas and then running three different load cases. Stiffness requirements were set to help ensure that material usage was only obviated in acceptable places.

In another instance, however, the approach was used later in the development cycle to optimize a rear axle to enable it to pass physical testing when only minor modifications were possible. Despite this restriction, stress reductions of 26% were achieved, which translated into a 233% increase in fatigue life.

Thellner says that Scania is still defining its use of these software tools. One factor that has to be overcome, he says, is getting people in various CAE operations to trust the tools and use them. In part that is about instilling appropriate confidence in engineers and in part about wider “organizational theory”; in other words, changing the way people work from one methodology to another.

Thellner's admission is a confirmation that, by itself, introducing sophisticated software tools is not necessarily sufficient to impact positively on a complex engineering process. The requirement also exists to ensure that people who will use the tools are confident and enthusiastic about doing so. He says that there are roughly half a dozen CAE groups within the company's product development organization. Use of the tools varies from complete adoption to initial exploration.

Seeking Global Test Validity

Another company that is exploiting specialized CAE software packages to simulate fatigue life in automobiles is Fiat Chrysler Automotive (FCA). In this case, some differences exist.

One is that the vehicles involved are passenger cars rather than trucks. A second is that the software supplier is a German company Siemens PLM and its Virtual Lab Durability and Virtual Lab NVH (Noise, Vibration and Harshness) analysis programs, along with its LMS Tecware tool for the manipulation and processing of test data. A third is that a crucial objective at FCA is not simply to reduce the need for physical testing of complete cars but to calibrate an equivalency between two test-track facilities—one in Italy and the one in the U.S.—to better ensure that test data generated in either location has global validity.

Fiat's aims are explained by Marco Spinelli, head of the chassis CAE department for FCA Italy, the European business unit that previously was known as Fiat Group Automobiles. He says that the combination of design software packages is now firmly established within the company to help predict road loads in the concept phase and as an aid to designers before a prototype is developed. But, in order to obtain an accurate load prediction, it is necessary to both validate sub systems, components and the full vehicle model virtually and record data from actual operational conditions on a test track.

Spinelli says that when a new car is developed, a huge amount oCircuito di Balocco test track near Turin, Italy. Image source: trackreviews.comf data is required to calculate the fatigue of different parts and components in order to ensure vehicle reliability. However, when new projects start, such detailed data typically is not available. Vehicle makers therefore often rely on a mix of past experience and measurements of predecessor vehicles that have a roughly equivalent mass. However, load data derived in this way may not be accurate enough because different vehicles possess different structural flexibilities and suspension systems. As a result, load transfer to different interface points will not be identical.

The solution, says Spinelli, is virtual simulation. The sophistication of “virtual drive" capabilities is now such that models of the car, road, driver and tires can be created and run together to obtain all the road loads that are required. Then, “when you have the loads and the material properties of all your components and parts, you can obtain fatigue maps which enable you to correctly set the dimension of your components and parts." This is accomplished with the Virtual Lab Durability package, which requires road load data. Spinelli says this is obtained either by direct measurement or through calculation with the aid of the MSC Adams/Car multibody software tool.

To the Track

Formal approval, though, still requires physical testing for which Spinelli says two options exist. The first is to “put your driver in the car and drive it over a million kilometers." The second is to “perform the fatigue experiment with increased loads correctly defined through equivalent damage in a laboratory or on a proving ground."

Fiat seeks to develop an equivalency between its EU and U.S. test tracks by analyzing a range of durability-specific characteristics. Image source: Fiat Chrysler Automotive For FCA, two proving grounds are involved. The first is the Circuito di Balocco test track near Turin, Italy. The second is a test track near Chelsea, Mich. The first has more than 65 kilometers of test tracks. The second offers more than 160 kilometers of on—and off-road surfaces and provides a test environment that is particularly appropriate for the demands of North America, which are more likely to include trips over rugged terrain.

Nevertheless, the ability to compare and correlate European and U.S. durability procedures is important for the company. As such, it uses LMS Tecware to develop an equivalency between the tracks by analyzing a range of durability-specific characteristics.

“We can get mission profiles for equivalent fatigue damage," says Spinelli. “If we can do that in different proving grounds, we can check and verify cars in different worldwide locations by using different profiles."

The end result is that both proving grounds are "fully equivalent and interchangeable." In practical terms, that means that there is no longer an EU or a U.S. proving ground. Instead, the company operates two “worldwide" facilities with design software as a shared asset.