Matt Greulach
ITI Manta

Because of their functionality and affordability, computer-aided software tools have forever changed the product development landscape. Today CAD, CAM, CAE is standard equipment for all companies regardless of their size or the industry they serve; yet as these tools continue to automate engineering, analysis, and manufacturing functions the need for hands-on data collection and field-testing remains strong. In fact testing is moving up the product development chain in its early stages where testing has a direct impact on product quality, time to market, and development cost.

History
Years ago product testing was performed almost exclusively to help pinpoint problem areas or to compare how a product stacked up against competitors. It was not until the 1960s that engineers began to think of testing in a different light. Like a lot of technologies, new testing resulted from industry funding of university research. The University of Cincinnati's Mechanical Engineering Department is one well-known example. In 1968 testing methods developed and enhanced while at the university stepped into the corporate world as a group of engineers, lead by Dr. Jason Lemon, left the university to form Structural Dynamics Research Corporation (SDRC). Many of these same pioneers, including Dr. Lemon, went on to start International TechneGroup Incorporated (ITI). Over the years, advanced testing methods have taken a more and more prominent, as well as increasingly important role in systems engineering and simulation driven product development.

Case Study: Volvo Graders
Volvo Motor Graders began manufacturing in 1875 under the Champion name. The company pioneered new equipment designs quickly developing a reputation for quality and reliability in more than 95 countries. With manufacturing centers throughout the United States, Canada, and Brazil, Volvo continues to provide a comprehensive line of industry-leading motor graders. Volvo has a long-standing reputation for quality. Each piece of equipment is engineered and manufactured to meet a wide variety of usage scenarios -- from frigid conditions to searing heat. So when a problem does arise with one of their graders in service, the company naturally has an urgent desire to correct the situation.

Weld Cracks
A small number of graders were experiencing cracked welds in the vicinity of the fuel tank when plowing snow. In order to correct the problem Volvo first needed to understand the root cause for the failure:

• Were the welds bad?
• Were material property vibrations a part of the problem?
• Was this the result of cold weather and low temperatures on weld and material properties?
• Were the dynamic loading conditions due to plowing, ice and snow?

For sure it was not a mere coincidence that Volvo would ignore.

Volvo enlisted ITI Manta to help them to understand and solve the problem. ITI Manta offers advanced mechanical engineering testing and analysis services that help industry to solve some of their most severe and costly engineering, design, and manufacturing issues.

From experience ITI Manta engineers knew that cracks usually can be traced to metal fatigue resulting from a combination of systems dynamic and complex dynamic loading and duty cycles. The Volvo / ITI Manta team began looking at modifications made to the grader by the operator; changes that would influence vibration and dynamic behavior.

The Solution

Working with engineers from Volvo, ITI Manta engineers designed a two-phase test and analysis plan to understand and resolve the problem. The initial test phase included instrumenting a grader and collecting operating data. Data was then analyzed to pinpoint specific grader dynamic behavior including vibration, as well as operational dynamic loading and duty cycle information.

Investigation of the problem began to identify and focus on dynamic behavior effects influenced by tire chains. In order to improve traction in the snow and ice, grader operators often installed tire chains. While this helped improve traction, it also directly impacted grader dynamic vibration behavior. These high vibration levels added higher stress levels and much higher cycle counts on some of the welds securing the fuel tank to the frame. In order to validate the root cause of the failure modes and design changes to resolve the issue, a carefully detailed test and analysis effort was implemented.

Using the strain and acceleration data that ITI Manta collected at key locations of a grader with chains installed on the rear wheels, and without chains installed, order tracking software was used to display the frequency content of the collected data. This helped determine both strain and vibration magnitudes, with and without chains, under various severe operating conditions.

Next grader vehicle dynamic behavior was ascertained using controlled force testing. In this case, impact modal vibration tests were performed to determine the natural frequencies and mode shapes (deflection patterns) of the entire vehicle but especially in and around the fuel tank region.

The analyzed data revealed areas where grader vehicle natural frequencies in the fuel tank area matched the high vibration and strain data amplitude content with chains installed. Test results proved conclusively that the weld failures were due to fatigue caused by tire chain dynamic loading conditions that excited grader vehicle natural frequencies with high vibration and strain on the fuel tank welds.

Figure 1 --- Operating Data --- Strain --- No Chains

Figure 2 --- Operating Data --- Strain --- Rear Chains



Figure 3 -- Strain Gage Amplitude

Testing revealed that adding chains to the tires had an adverse effect on vibration and dynamic strain levels as is shown in Figures 1 and 2. Figure 1 shows a strain gage that was placed on the fuel tank near the area where structural cracks occurred in the field for the condition of no chains. Figure 2 shows the same strain gage for the condition of chains on the rear wheels. The plots show a gradual acceleration from slow speed to maximum speed, and then a gradual deceleration to slow speed again.

ITI Manta then utilized order tracking software to analyze the frequency content of the strain gage data during the gradual acceleration from slow speed to maximum speed. The order tracked plot of the strain gage data for an operating run with chains on rear is shown in Figure 3.

Figure 3 displays the strain gage amplitude in the form of a Campbell diagram. The increasing colors indicate increasing strain amplitude. Typically, the diagonal lines of a Campbell diagram show the orders of a rotating system. The vertical lines can indicate system natural frequencies.

In the Figure 3, the first diagonal line to the left is approximately 9th order of the rear wheel speed. The "9" corresponds to the number of chain strips per revolution of the wheel. A vertical line pattern of increased amplitude occurs around the 25 Hz area.

The controlled force impact modal structural dynamic tests were performed on the motor grader with focus on the fuel tank area revealed a natural frequency at 24.5 Hz with the mode shape being a panel mode of the vertical fuel tank wall. This 24.5 Hz mode correlates precisely with the high amplitudes seen in the Campbell diagram of the operating test in the 25 Hz region. The mode shape shows a large amount of relative motion in the panel with related high strains occurring precisely in the areas of weld failures.

Solving the Problem

With this knowledge, understanding, and validation of root causes and failures, dynamic loading conditions, grader dynamic behavior, and validation of vehicle dynamic modal models and finite element models in the fuel tank region, it was not difficult for Volvo Grader engineers to incorporate modifications to the structure to eliminate the problem.

The structural modifications shifted the problem natural frequencies out of the range of tire chain dynamic excitation frequencies and lowered strain levels in the fuel tank mount region significantly.

Conclusion

Like other product development processes, testing has steadily progressed in terms of sophistication and effectiveness. Today data collection equipment has become more portable, compact, and effective allowing tests to be performed on virtually any product in all types of conditions and environments.

Most products are becoming increasingly complex. Usage is becoming more varied and in many cases it is difficult for product designers to know their products and their competitors' products are being used.

Today test data is being used not only to fix problems -- but more important, to identify usage profile information, competitive product information and to validate increasingly complex product, system, sub-system, component simulation models used in upfront system engineering and analysis leads design efforts. Without increasingly advanced testing capabilities and methods, new product development processes and implementation methods providing significantly faster time to market and development productivity are not possible.

The old design, build, "shake and break" methods simply will not provide customer expectations in terms of product performance, quality, cost, durability, and reliability in competitive time nor at competitive cost.

About ITI Manta

ITI Manta provides advanced testing and analysis services for most discrete part manufacturing industries. ITI Manta is a part of International TechneGroup Incorporated (ITI).

ITI is a global leader in the new processes, implementation methods, integration, infrastructure and global collaborative capabilities required to deliver "Breakthrough Improvement?" as measured by time to market and development productivity needed by clients to deliver products with world class performance, quality, cost, and reliability to global markets.