The three major design components of this project are the (1) actuation mechanism (ie. what's providing the pushing/pulling force ), (2) guided linear motion (ie. what's providing smooth motion in one degree of freedom), and the (3) locking mechanism (ie. what is going to hold the foot in place once proper alignment is achieved. Most integral to this team’s responsibility to Solar Turbines is the investigation and selection of a viable linear actuation mechanism that is capable of achieving high degrees of accuracy and precision.
The preferred configuration of each of these components are as follows: (1) Jacking bolts & worm gear screw jack, (2) sliding contact bearing, and (3) bolt-to-frame locking
Provides the necessary pushing/pulling force to position the strut foot with high degree of accuracy
Sliding Contact Bearing will provide linear motion of the strut foot in one degree of freedom
Locking bolts will provide the necessary clamping force to hold the foot in place while the gas turbine is in operation
Most integral to this team’s responsibility to Solar Turbines is the investigation and selection of a viable linear actuation mechanism that can be deployed to the field on their new T-250 six-strut mount design package. The early stages of this project consisted of deep research in the best solution for actuation. Several designs were weighed against each other according to the following criteria: ease of alignment, speed, accuracy/precision, reliability, envelop, and cost. A summary of our findings is presented in this Pugh chart. As seen in the chart, the jacking bolt and the worm gear screw jack scored the best and were both considered as final solutions for the actuation mechanism.
The first solution to Solar Turbines is the jacking bolt. A base serves as a linear guide for the foot to slide on while two opposite facing jacking bolts adjust the position of the foot through the input of a torque wrench.
The chosen jacking bolt for actuation is ¾” in diameter with a tenth of an inch pitch. The jacking bolt is highly accurate and boasts a maximum required torque of 36 foot pounds to operate.
With the use of a dial gauge this solution can achieve accuracy of 1 thousands of an inch, 50 times better than the required accuracy. This design is by far the least complex and expensive. It has a low weight and profile making it an ideal candidate for an actuation system.
We were constrained in our design by a necessary 2 inch maximum extension of the jacking bolt, and a 30lb input force using a 2ft lever. We had to then optimize our choice of jacking bolt to meet those requirements. From the models above we were able to identify that optimizing the diameter of the jacking bolt would be much more valuable than the pitch since it is more closely related the necessary input force.
The necessary diameter of the jacking bolt to obtain the required buckling strength was examined. The model shows the minimum diameter of a jacking bolt for a 2 inch maximum extension. This model helped us establish the desired diameter of the jacking bolt.
Choosing the right jacking bolt to account for the necessary input torque was a challenge, but we developed models to examine the input torque requirements for different pitch and diameter jacking bolts, and the buckling loads for each. We landed on ¾” extreme strength bolts being the optimal choice for our design. These bolts give us a maximum buckling load of almost 70,000lb utilizing the exponential nature of the buckling load function with diameter. This gives us a safety factor of 7.4 for the loading conditions during alignment. The final required input (tangential) force from the technician on a wrench was measured through testing to be 22lb (matching out 21lb estimate from the above model) for dry sliding friction. With the use of grease this number was found to be about 16lb, well below the 30lb requirement.
Construction consists of a 36:1 gear reduction between an input worm to a worm gear which is capable of producing a high torque and thus a maximum thrusting capacity of 65kN. The worm gear will convert rotational to linear motion via a ball screw with a 50 mm diameter and 10 mm pitch. In order to push the worst case load of 9448 lbf, the required drive torque applied to input worm shaft is 4.1 lbf-ft. The most cost effective and user-friendly approach to driving the screw jack is through the use of a hand crank of 5 in. radius. With this configuration, a technician will need to apply only 10 lbf of tangential force to the crank. Every full rotation of the hand crank results in 0.011 in of linear displacement of the screw making the displacement of the foot very controllable and predictable. The screw jack will be attached to the foot via a clevis joint to allow pushing and pulling from one side of the foot.
At its core, the worm gear screw jack is an improved model of a jacking bolt in that it utilizes a large gear reduction that converts high rotational speed into a high torque output. Worm drive is non-reversible, meaning that the driven worm gear cannot drive the worm. This condition offers more security to the system as it will reject unwanted rotation from the screw drive itself and in the event of foot slippage, the worm gear jack screw will not backdrive.
Locking the Foot Down
Locking the feet in place keeps the turbine in alignment, and keeps the turbine safe during operation, and in extreme situations such as shipping and offshore rig deployment.
Locking the foot will be accomplished with (6) 3/4"-10 bolts, high-strength steel insert locknuts, Belleville Disc Springs, and a thread locking compound. The combination of these measures will keep the bolts from loosening under the vibrational load of the turbine's operation.
The bolts will provide 30,100 lbs of clamp force each, for a total of 180,600 lbs.
Due to overall scale of this project and financial restrictions, a full scaled model was not produced. However in order to test the how the effects of the stick-slip phenomena will hinder our design's positioning capabilities, a scaled down model was produced. Due to its accessibility, the jacking bolt solution was chosen as the actuation mechanism for our test. In order to simulate the 12,000 lbf loading that the aft strut will experience, a hydraulic press was utilized to apply extreme vertical loading. To simulate the horizontal loading that the strut will experience in reality, the vertical load from the hydraulic press was increased.
The cause of stick-slip:
Difference between Static and Dynamic coefficients of friction.
The pushing force required to break static friction is higher than the force required to move during dynamic friction resulting in a step-rise in acceleration and thus potentially imprecise adjustments of the foot
The test rig was designed to increase stiffness and thus decrease the effects of stick-slip. In order to simulate the load moving with the foot, 3 steel rollers were placed in between the press and the foot. To measure the total displacement of the foot and to characterize its motion profile, a linear displacement dial indicator was placed at one edge of the foot. During testing, slow-motion video was taken of the dial indicator to see if the stick-slip phenomena were significant in our system.
Videos demonstrating the use of the testing set up by Team Member Vicente and feedback from the dial indicator
Multiple tests were ran with with various loads applied by the hydraulic press; also, assorted combinations of greased vs non-greased surfaces were applied on the threads and sliding contact surfaces of the test rig. A 20 inch wrench was used to rotate the jacking bolts. After observing and recording the required input torque, the distance of foot travel, as well as the motion profile for each test it was determined that stick slip would not be an issue. Specifically, under the worst conditions the required input torque was still less than 45 lbft. The movement achieved could be described as like cutting a hot knife through butter, it was both smooth and consistent. As a result, precise movements in increments of 1 thou were easily achieved during testing with maximal load. Equally as important and worth mentioning was that our experimental results validated our torque and precision estimations.
Simulated kinematics with 12000 lbf external component force
Actuation modeled as spatial contact force between screw and foot geometry
Contact Characterization:
Stiffness [N/m]
Damping [N/(m/s)]
Transition Region [m]
The foot mount was modeled using ANSYS with maximum loading during alignment of 48,000 lb. The result was a maximum deflection of 1.5e-5 in. This analysis has helped verify the performance of the foot mount and that the deflection will remain within the allowable tolerances.
The pushing and pulling of a worm drive on the foot was also modeled. The result was a maximum deflection of 5.082e-4 in for pushing and 1.022e-3 in for pulling. This results are within the allowable tolerances.
Analysis was done on the deformation of the stiction test rig to evaluate the force it can be subjugated to during testing. The maximum deflection on the pushing wall of the foot was found to be 9.315e-4 inches
The foot's natural resonance frequency's were estimated using SolidWorks. Avoiding resonant frequencies will ensure the foot will not oscillate uncontrollably and jeopardize the integrity of the foot mount. It is difficult to gauge the frequency and amplitude of vibration each strut will be subject to, but as the six-strut mount design matures through the design phases, these are the natural frequencies to avoid during engine operation.
FBD of foot during alignment (Pushing)
FBD of foot during alignment (Pulling)
FBD of foot during Operation
The binding ratio of a linear rail system was examined to determine the forces the foot would encounter if there was an offset distance between the applied force and the center of the rail. These forces could create added friction or bind/lock the foot on the rail.
During alignment, our friction coefficient is μ=0.6 and length of the foot is L1=19.5 inches, so maximum allowable lever distance between the applied force and the center of the foot was found to be D1max=17.91 inches .