Final Design

Overview

The fatigue tester is based on a Scotch Yoke mechanism, in which the rotary motion of the motor is converted to linear motion via a track roller and a linearly constrained guide. Five major components were designed in order to implement such a mechanism and fulfill the requirements of this project: Timing pulley, Yoke, Guide, Spring Set, and Test bed. CAD and the final build can be found in Figure 2. Also found in Figure 2 is a flowchart of the motion path, which describes how the motion of the motor reaches the sample.

Figure 1: Scotch Yoke motion. Circular motion of a rotating

track roller pushes against a linearly constrained guide. By

increasing the eccentricity of the track roller, the stroke of

the guide also increases.

Figure 2: Full build of fatigue tester

Timing Belt System

The unused motor provided by ATA is a U.S. Varidrive motor that has a built-in variable speed adjuster that can deliver 101-705 RPM to the output shaft. From this output shaft, a timing belt transmits the rotary motion to a secondary shaft that is free to rotate. Since the motor rests near the ground, the main purpose of the timing belt is to lift the system off the ground and prevent the need to crouch while using the fatigue tester. The timing belt system uses a 3048-8MGT-30 PowerGrip GT2 belt that matches the motor spec of 705 RPM and 3.7 kW (5 HP).

Yoke System

The Yoke system is responsible for converting the rotary motion of the motor to linear motion via a Scotch Yoke mechanism, which is illustrated by Figure 1. An overview of each component of the Yoke is given in Figure 3. The Yoke consists of a track roller, which is both threaded onto a rod and rotationally constrained by a steel housing. As the threaded rod is turned, the Yoke moves linearly and the eccentricity increases, as shown in Figure 4. The threaded rod can be turned via a hand dial, which displays the rotations with a 100 increment per rotation resolution. This hand dial allows the user to accurately define the stroke of the tests in a repeatable manner. To avoid undesired motion of the system, counterbalance plates were added to the side of the housing to account for the non-symmetry system when the eccentricity is non-zero. 

The Yoke was designed to endure 22.3 kN tests without fatiguing. SolidWorks static stress analysis was performed on the system and found that the maximum stress was 187 Mpa, which occurred at the interface between the Yoke and the rotary shaft. The housing was made out of hardender 17-4 steel with an endurance limit of 466 Mpa. 

Figure 3: Overview of Yoke system with key components

Figure 4: Changing eccentricity of Yoke roller and resulting circular motion

Guide System

Constrained by linear bearings, the guide is in direct contact with the Yoke and is responsible for the linear motion of the fatigue tester (Figure 4). The guide is connected to the main I-Beam tower via 8020 to mitigate unwanted bending strain in the framework while increasing the overall system stiffness. An adjustment nut allows the spring set and load pin attached to the output end of the guide shaft to be modified and accommodate samples of varying thicknesses. A spring at the top of the subassembly bears the free weight of the guide shaft, preserving the accuracy of tests conducted at low load ranges (where the weight of the guide would exert a non-negligible force on the specimen relative to the applied load). Ball transfers are to be adjusted and locked in a position such that they touch off on either side of the guide, such that the guide’s rotation is constrained during testing. Additionally, the installment of the spring set prevents backlash by maintaining compressive pressure between the guide and the specimen being tested (assuming contact between the load pin and specimen is maintained).

The guide is design to withstand up to 1x107 cycles of 22.3 kN load tests without fatiguing. The hardness of the guide material that interfaces with the Yoke is sufficient to withstand the Hertzian contact stress that is develop, as specified by the manufacturer of the Yoke track roller. Additionally, the linear bearings were chosen to survive the radial loads during max load tests.

Figure 4: Overview of the upper portion of the guide system

Also attached to the guide is an adjustment nut, which lies above the spring set and allows the load pin to be moved up and down to account for different sample thicknesses. Additionally, this adjustment nut can be used to preload to springs and set the minimum force that the sample will feel during the test. Figure 5 illustrates the functionality of the adjustment nut.

Figure 5: Functionality of the adjustment nut to compensate

for varying sample thicknesses and preloading the sample

Spring Set System

The spring set consists of five radially spaced springs that couple the guide to the loading pin. Displacements of the guide are transferred directly to the springs and the resulting force observed by the sample is determined by Hooke’s law. As a result, the sample force is no longer dependent on the stiffness of the material and relatively large displacements are still necessary to apply loads to very stiff material. This provides better resolution when adjusting the Yoke for a desired force. Each spring rests between the two metal spring plates within pockets that impede the horizontal motion of the spring. These pockets have two depths, which allow for different dimensioned springs for different eccentricity resolutions. Within each spring is an acetal plug that is dimensioned for a tight fir within the spring. This fully couples the guide rod to the load pin and helps mitigate the buckling potential of the springs. Figure 6 provides a cutaway image of the spring system. 

To obtain better resolution when adjusting eccentricity, five spring sets were chosen, each with a designated operating range within the required 0-5000 lbf. Table 1 provides a summary of the five sets chosen. Each spring was analyzed for resonance, fatigue, and buckling and was determined to be sufficient for the loads and deflections within its operating range.

Figure 6: Overview of spring set design

Table 1: Spring sets spring constant (per spring), operating range, and largest deflection within the operating range.

Test Bed

the test bed’s main function is to hold the sample for testing and allowing testing for different sized samples. In holding the samples, the test bed does not interfere with the bending test in anyway. The testbed was designed for economy, using 8020 to develop a stiff structure at a low cost. Its primary features are the sample support stanchions, the centering link, and the body of the structure (Figure 7). The centering links provide mirrored motion of the stanchions to accommodate different length samples. Analysis of the test bed confirmed its capacity to withstand a 22.3 kN test. The maximum stress was found to be 62 Mpa, which is well below the endurance limit of aluminum (given 1e7 cycles) of 158 Mpa.

Figure 7: Overview of test bed and its design features

Base Plate

The baseplate rigidly connects the motor to the test fixture through 8020 extrusions and aluminum I-beams. Due to the assembly’s large size, resonant frequencies proved to be a high risk factor of the baseplate. Close-out and shear plates were added throughout the system to increase the rigidity and resulting resonance modes of the system. Figure 8 shows the FEA results of the modal analysis performed on the system. These simulations assumed the base plate was bolted to the ground, which is crucial to avoid resonance near the 18 Hz operating frequency. The first mode of the system was found to be 29.3 Hz. Additionally, static simulations revealed the max stress to be 58 Mpa, which is located on the aluminum plate at the interface between the test bed and the base plate.

Figure 8: FEA modal results for the fatigue tester system. First

mode of the system was found to be 29.3 Hz.