Supplier Spotlight

Show Reviews

Automotive Testing Expo Europe 2017 Show Review

Click here to read

 

Automotive Testing Expo Europe 2016 Show Review

Click here to read

 

Automotive Testing Expo Europe 2015 Show Review

Click here to read

 

Automotive Testing Expo North America 2014 Show Review

Click here to read

 

Automotive Testing Expo India 2014 Show Review

Click here to read

 

Latest Video


Euro NCAP crash footage

LV

Range Rover Velar crash test montage October 2017

Click here to watch the video


Accelerate acoustic troubleshooting

LV

A modular digital microphone array from Siemens PLM Software gives an instant overview of sound sources on any noise-generating object

Click here to watch the video

If you could pick only one, where would you prefer to perform climatic testing?

Articles

« back to listing

MTS Systems Corp: Hybrid simulation

MTS Systems has developed an iterative hybrid simulation technique for complex ride comfort and durability applications

 

Today’s vehicle developers must validate new technologies on vehicle platforms with numerous variants, while using fewer resources and meeting tighter deadlines. Hybrid simulation — the integration of physical tests and virtual models — holds tremendous potential to accelerate and enhance traditional development processes, but it presents considerable technical challenges.

One of the most daunting challenges is when the physical test rig and the virtual model can’t interact in real-time. To enable hybrid simulations in situations where real-time techniques are not possible, MTS Systems Corporation has developed Hybrid System Response Convergence (HSRC). This innovative technique makes hybrid simulation more practical and efficient for a wider range of applications throughout the automotive industry.

Advantages of hybrid simulation
Vehicle manufacturers worldwide are beginning to appreciate how hybrid simulation can help validate components and subsystems earlier in the development cycle. Prior to hybrid simulation, validation of components and subsystems had to be done using one of three methods — none of which was ideal. Tests could be configured to apply loads gleaned from measurements of real parts and vehicle prototypes, however, these are only available relatively late in the development process so any problems discovered are expensive to resolve. Alternatively, test loads could be adapted from similar vehicles or platforms, but this compromises the accuracy of tests, which ultimately affects the quality of component design. Test loads could also be predicted from a completely virtual environment, but this method is even less precise than adaptation.

Hybrid simulation solves this problem. Because it combines the physical with the virtual, it allows OEMs to configure lab tests for subsystems and components much earlier, effectively compressing the development cycle. OEMs can develop accurate lab tests for subsystems and components without field data acquisition, as well as understand and improve CAE model correlation. They can also validate test specimens without “force fitting” or predicting loads.

MTS has developed and pursued two distinct implementations of hybrid simulation for ground vehicles that integrate mechanical test systems: Mechanical Hardware-in-the-Loop (mHIL), and Hybrid System Response Convergence (HSRC). While mHIL occurs in real-time, HSRC is an iterative process that develops system inputs until compatible dynamic behavior converges between the physical and virtual components. The results that HSRC produces, however, are dynamically equivalent to a real-time hybrid simulation.

How HSRC works
A typical HSRC application consists of a physical test vehicle (body and suspension) combined with a set of virtual tires, all of which is “driven” over a digital, three-dimensional test track. The test vehicle is installed on a road simulator and can be excited at the vehicle spindles, which act as the interface between the real vehicle and the virtual tires. The road simulator is equipped with force and motion sensors (typically 6DOF) to measure the spindle dynamic behavior during each iteration. On the virtual side, the tires are maneuvered by a virtual “driver” over the road.

The system generates results through sequential execution of the physical and virtual simulations. An initial excitation sequence is played into the physical vehicle and the resulting spindle forces and motions are captured. For each control axis, one of the spindle responses (force or motion) is used to control tire behavior in the virtual simulation. For example, the vertical force measured on the physical rig can be used to control the vertical loading of the virtual tire as it is maneuvered over the virtual road. This creates a unique vertical motion of the virtual tire center, based on the tire dynamics and the road profile.

The HSRC system then compares the rig motion that caused the vehicle force with the tire motion measured when the same force was applied to the virtual tire. If the vertical motions are not equal, the two halves of the hybrid system are not yet operating with dynamic compatibility.

Using the difference in motion, or convergence error, the HSRC control model changes the dynamics of the physical rig, resulting in new force and motion measurements. This creates a new virtual test condition and new convergence error results. Through successive iterations of this process, HSRC ultimately creates a road simulator drive file in which the measured vertical force, when applied to the virtual tire, results in the same motion that generated the measured force on the rig. At this point, the two halves of the simulation have converged and are dynamically compatible. In other words, they are behaving exactly as they would if they were coupled together as a real-time dynamic system.

While the simplified example above describes the convergence of only vertical motion at a single spindle, in an actual HSRC simulation convergence would occur for all degrees of freedom simultaneously, and the final drive file would represent converged dynamic behavior at all four vehicle spindles.

When and why HSRC makes sense
As an iterative process, HSRC takes longer to realize a simulation than real-time techniques, such as mHIL. So why choose HSRC over real-time? The short answer is complexity.

Real-time hybrid simulation techniques require the ability to do all of the following: 1) simulate virtual elements in real-time, with no additional CPU time required; 2) control the physical test in real-time without pre-programming, via perfect tracking control; 3) couple all system elements with a continuous high-speed data link, with no transport delay.

Here’s where the problems arise. The constraints of real-time hybrid simulation are typically incompatible with common vehicle simulation requirements. There are three reasons for this.

First, simplified virtual vehicle and tire models that can run in real-time are inadequate to represent the more complex vehicle responses needed for useful ride comfort and durability simulations. Second, durability test systems required to load the physical components are often too complex to deliver perfect tracking control. Finally, for these high bandwidth multi-axial test rigs, any transport delay in the system will destabilize closed loop behavior and compromise the integrity of the hybrid simulation.

Because of these limitations, it is not feasible to configure a real-time hybrid simulation that has the component detail and operating complexity necessary to perform most ride comfort and durability simulations. But the pressure to accelerate development remains intense, which means OEMs need a way to get real-time results without real-time hybrid simulation techniques.

That’s where HSRC comes in. It is ideal for hybrid simulations that can’t be executed in real-time due to the control requirements of the physical test system or the complexity of the virtual model (for example, models that involve elastomeric components and/or tires that exhibit nonlinear behavior). HSRC also provides logistical flexibility. Because the physical and virtual worlds are not interacting in real-time, they can be located in different labs.

Challenges of HSRC
While simple in concept, HSRC is difficult to execute correctly. To make it a viable and efficient simulation technique MTS had to overcome numerous challenges, both technical and operational in nature.

Among the key technical challenges was the problem of determining where to start simulations. To solve this problem, MTS developed a proprietary strategy for initiating the iterative process from a valid starting point, significantly reducing the number of iterations required for convergence. A valid starting point is also critical for avoiding inappropriate initial loads, which could easily damage the physical test article. Other technical challenges included devising a way to “drive” four “disembodied” (virtual) wheels and tires on a virtual test track in a coordinated way, despite the fact they are not connected to each other, as well as developing a compensation function to convert error into directionally correct drive file updates, enabling the simulation to proceed continuously toward convergence.

To make HSRC an operationally practical solution for developers, MTS integrated all of the solution components to minimize complexity for users. A number of activities, including coordinate transformations, DOF matching, polarity matching, virtual driving and process initiation, were designed to be managed “under the hood,” minimizing the risk of operator error and small mistakes that could severely compromise simulation quality and test schedules.

In addition, MTS developed a menu-driven approach to HSRC familiar to any engineer using RPC Pro software. This familiarity helps operators know what information they need to keep the simulation running, and reduces the need for extensive training.

Proven results
HSRC is an innovative technique for developing laboratory full-vehicle simulations, and it has been validated in six vehicle programs at five major OEM sites. These evaluations included a variety of vehicle configurations, tire models and digital roads. Convergence was successful in all cases.

Proof of concept for HSRC occurred with Audi in 2010. An Audi A5 Coupé was installed on a spindle-coupled road simulator (a 6DOF rig with 6DOF control inputs) combined with virtual vehicle tires (modeled using FTire) and Audi digital road sections running in a standard ADAMs simulation environment. Working with Audi, MTS developed a complete set of 20-channel control signals for the road simulator for three rough-road durability test profiles that represented select Audi proving ground surfaces.

In this initial evaluation, the physical spindle loads developed by the hybrid simulation were compared to vehicle spindle loads obtained from three other methods: road load measurement for the A5; road load measurement for a similar vehicle; and analytically predicted loads from a virtual vehicle model. Comparisons indicated that the laboratory vehicle test loads created with HSRC closely correlated to fatigue-critical loads from a physical road measurement, resulting in more appropriate loads than those predicted through analysis.

In the years that followed, MTS developed a number of enhancements to the HSRC — all of which improve the solution’s efficiency and broaden its scope of potential applications.

MTS worked directly with FTire to integrate tire simulations in a stand-alone application (outside the ADAMs environment), which improved the efficiency of iterative processing by 85 percent. This in turn enabled Chrysler to simulate more segments of road data in the same amount of time.

MTS worked with a major OEM to support a second industry-standard tire model from TNO, proving that HSRC can work with multiple tire models. The OEM evaluated different behavior on different tire models and then compared the results. Both tire models were evaluated and validated successfully with the same setup.

Another enhancement added a full track simulation. Working again with Chrysler and using an HSRC implementation involving a 4DOF spindle-coupled road simulator, MTS helped the OEM simulate the entire proving ground track instead of individual sections. This enhancement depended on Chrysler’s willingness to invest in a 3D topological scan of the entire track.

Pushing toward the future
Innovations in HSRC continue today. Currently, MTS is adapting the HSRC solution to create a technique for using fixed-body axle test configurations instead of the typical floating body. In this implementation, a virtual vehicle body, tires and wheels are used to test a real suspension system, enabling accurate testing as soon as the components are available — well in advance of the availability of a full vehicle prototype. MTS is also developing a technique to test a full vehicle with body restraints (both virtual and physical), which enables dynamic braking and cornering simulations. The challenge with both techniques is in correctly modeling the influence of suspension loads on body dynamic behavior.

The requirements for HSRC are straightforward. OEMs that want to adopt HSRC need to have RPC Pro software and an MTS 329 road simulator using a FlexTest controller. MTS is ready to implement HSRC for many other applications too, including other spindle-coupled road simulators as well as for other system applications such as engine mounts, steering systems, exhaust systems and more.

Contact
Dave Fricke, MTS Systems Corporation
Tel: +1 952 937 4000
Email: dave.fricke@mts.com
Web: www.mts.com

 

Read Latest Issue
Read Latest Issue

Web Exclusives

The importance of software engineering

Tier 1 supplier GKN’s software and electronics manager, Michael Schomisch, discusses how software development is becoming increasingly important in vehicle testing
Click here to read more



Daimler reveals how the GLC F-Cell went from testing to production

With the Mercedes GLC F-Cell SUV unveiled at Frankfurt this week, Daimler has revealed the extensive testing process the first production plug-in hydrogen-powered car was put through
Click here to read more



Speaker Q&A: Jerry Beeney, FLIR Systems

The business development manager, who will be presenting at Automotive Testing Expo in Novi, Michigan, discusses current measurement tools and techniques used to measure high-speed targets, including thermal cameras and contact and non-contact temperature measurement systems.
Click here to read more



Conti celebrates 50th anniversary of the Contidrom

The German tire giant looks back on 50 years of testing and development at its flagship facility
Click here to read more



Cruden launches new simulator technology

The system designed for ADAS and autonomous vehicle testing enables engineers to add as much hardware- or software-in-the-loop as they wish
Click here to read more




Supplier Spotlight

Supplier SpotlightClick here for listings and information on leading suppliers covering all aspects of the automotive testing industry. Want to see your company included? Contact jason.sullivan@ukimediaevents.com for more details.

فروشگاه اینترنتی فروشگاه اینترنتی سیستم همکاری در فروش ماهواره آنلاین اندروید کانال تلگرام چت روم دانلود فیلم فروشگظ;ه اینترنتی

Submit your industry opinion

Industry BlogDo you have an opinion you'd like to share with the automotive testing community? Good or bad, we'd like to hear your views and opinions on the leading issues shaping the industry. Share your comments by sending up to 500 words to john.thornton@ukimediaevents.com

Advertising

Recruitment AdTo receive information on booking an advertising banner please email jason.sullivan@ukimediaevents.com