Inconel Tubing Pre-Pullout
|Team Name||Lucky 13|
|Duration||Fall 2017 - Spring 2018|
The goal of the project is to provide the Boeing Company with a more ergonomic and consistent way to provide tubing material grinding prior to pullout. The purpose of the pullout is to join tubing in the most efficient and strongest way possible. This project aims to reduce the time and effort for the Boeing operators to obtain a successful pullout.
- 1 Problem Explanation
- 2 Engineering Requirements
- 3 Project Learning
- 4 The Design
- 4.1 Visual Feedback Subsystem
- 4.2 Grinder Fixture Subsystem
- 4.3 Force Feedback Subsystem
- 5 Testing & Validation
- 6 Documents
- 7 The Team
The Boeing Company's Tube, Duct, and Reservoir Center (TDRC) has asked the University of Idaho for assistance in creating a solution in their tubing pullout process. A tubing pullout is an alternate way to join two tubes in a duct system. This process is faster and stronger than other methods available. Currently, the cutting of the tubing is done by either a water jet or a laser cutter. Both of this methods create areas of weaker material around the hole that needs to be removed before the pullout can be done. This material is removed by grinding, using a soft stone, and is done by using a handheld grinder. The method requires the operator to hold both the tube and the tool. The completion of grinding is decided by the individual operator which creates inconsistency with time and pullout success. Not removing enough material will cause the pullout to tear and removing too much is a waste of time. The Inconel Tubing Pre-Pullout team aims to create a viable and easy to implement solution for Boeing's TDRC.
1. Reduce the time of the grinding process to less than 20 minutes
2. Solution will be justified with a business case
3. Operator will have 1 point of contact or less with tool or part
4. Discover an accurate way to measure ground surface
5. Provide amount of material to be removed based on type of material (TBD using tests)
6. Solution must accommodate 15" duct section
Activities and information the team gained from the project learning phase of this project.
Initial Client Meeting
The team went to Boeing's TDRC in Auburn, Washington at the beginning of the project to have a face to face meeting with the client and witness the pullout grinding process first hand. Some of the things we learned from the meeting were:
- Every operator likes to hold the part and grinder a different way
- Whether the pulling die is hot or cold depends on the material
- Inconel and Steel require a cold pull
- Aluminum and Titanium require a hot pull
- The water jet is preferred to the laser when cutting the pullout holes
- There is use of both Inconel 718 and 625
- Pullout takes less time to manufacture than a fish-mouth joint and has a higher strength after weld
- Boeing is primarily concerned with operator's wrist position and use during grinding
- Boeing would like to help avoid stress and overuse injuries
- In addition to grinding, some larger pullouts also require polishing
|Material||Inconel 625||Inconel 718||304 Stainless Steel||6061 Aluminum||3AL-2.5V Titanium|
|Ultimate Tensile Strength (MPa)||880||1375||505||310||620|
|Yield Strength (MPa)||460||1100||215||276||500|
Initial Design Ideas
|This idea was a computer controlled grinder holder fixture that would be completely automated. The grinder would be able to move in the cylindrical coordinate system to be able to grind all services of the tubing's complex curves. We decided against this idea because it was more complicated than was required in the scope of this project and would have required more time than we had to make it work correctly. The client shared similar concerns.|
|This idea was a fixture that would be attached to the grinder that could measure the force being applied to the grinding surface. This force measurement would then be read out to a digital display for the operator to see. This idea failed to meet the engineering requirements for ergonomics but, we liked the idea of the force feedback system.|
|This idea was a flexible arm that could hold the tubing as the operator grinds on it. This idea was ultimately scrapped in favor of holding the grinder because of the wide variety of tubing sizes that need to be accommodated.|
We decided on two different ideas that would be complementary to each other. The first is a visual inspection of the ground surface and the second is an arm to fix the grinder to the table with a force feedback system with a digital display. Through further investigation of the problem and the time and resources available to us, we have to decided to remove the force feedback system and concentrate our attention on the grinder arm fixture and the visual feedback system.
Visual Feedback Subsystem
The next step for this subsystem will be to construct the microscope assembly and test it on both ground and not yet ground parts. Then we would like to create a system to etch or mark the lip, so that operators have a visual representation of the material they need to remove, approximately 5 thousandths of an inch. This will further accomplish the process being quicker and more consistent.
The tube holds a magnification lens on one end and a light on the other, with a slot for the edge of the part in between the two. This design must be changed, as compound magnification will be required, and the tube must have a smaller diameter to fit into all the necessary slot sizes. A proof of concept was built using a toilet paper roll, flashlight, and 2X magnifying glass and we were able to see large cracks. Modification to this system is required, but the basic concept will remain the same.
Visual Feedback Prototype
The prototype design for the visual feedback system will focus on the use of a 50X magnification pocket microscope. The grinding range that will need to be visible is approximately 5 thousandth of an inch, or .127 mm at 50X magnification; this range will appear as 6.35 mm when viewed through the microscope. The end of the microscope will have a fixture, with a slot for the part, and which holds the back light. The fixture to hold the system may be machined out of PVC or aluminum or 3D printed depending on the complexity of the finally design and the strength and durability required. These factors will be determined later in the design process.We have completed a visual inspection prototype, made of a pocket microscope and 3D printed parts, and we have gathered useful observations, using this 50 times magnification tool.
We briefly considered etching or marking the edge to show the desired grinding distance. Etching wouldn’t work because it would introduce more stress concentrations. The Idea of marking was discarded as unfeasible, because there are too many variables, the distance that needs to be ground is unknown, the consistency of the mark being made, and the geometry of the ground edge is unknown.
Visual Feedback Station
We will now integrate an inspection station for the workers, which will use an all in one digital microscope to view between 20 and 400 times magnification. This tool will provide an easy and fast inspection for operators, as well as being a useful tool for future experimentation.
Grinder Fixture Subsystem
This system will provide the support for the grinder and will house the force feedback system. Because it will be used in a manufacturing setting, there will be a large factor of safety to protect against impulse loads regularly found from accidental contact with the machine. The main goals of this machine are to increase consistency and durability, and insure ergonomic use to lower workplace repetition injury.
The initial sketch for the grinder fixture has the grinder at an angle but it will be fixed vertically with the grinder tip pointing down. We have also removed the camera microscope in favor of the handheld visual inspection tool. The grinder for the system is the Bosch DG355LCE.
First Model Iteration
The grinder sub-assembly is essentially two pieces, an arm that telescopes up and down and the fixture that will hold the grinder in place. The arm will be made from heavy duty pipe with a flange on the bottom to fix the arm to the workbench. The upper part of the arm will be smaller so that it can fit inside the base. This will allow for the ability to telescope the arm. Telescoping the arm will allow the operator to move the grinder tip into the best position for the duct size being ground. The grinder fixture, the piece that will actually hold the grinder, will be made from aluminum. It will have a similar design to the V-blocks used to hold cylinders for machining. This will allow different kinds of grinders to be used in the same fixture. The fixture will be tightened into place using wingnuts. The sensors for the force feedback system will be placed somewhere on the arm to measure moment or torque, which can be calculated into force applied at the stone tip. The design for the telescoping arm will likely not change for the next iteration. The grinder fixture needs a redesign for the optimization of the force feedback system. For more information about the force feedback system, refer to that section.
The final design for the arm has not significantly since the first iteration, but we have added features that will allow for better ergonomics and ease of use for the operator as well as ease of manufacturing and lower cost. We have since sourced and obtained a grinder that will be used in our final design and will be given to our client with our solution, a Bosch DG355LCE die grinder. It is a large grinder with variable speed and the ability to survive in the harsh conditions of a manufacturing environment. The grinder will be mounted to the pipe arm using a fixture that can support the grinder both vertically and horizontally. The grinder "hand" will plug into the arm using a 2" nom sch. 40 pipe section similar to the way a trailer hitch works. This will allow our client to easily fabricate its own grinder hand for additional grinders. The hitch pin will allow for easy removal and installation, satisfying our ergonomic requirements.
Because our client must be able to grind on tubing sections between 1"-8", we have given the fixture the ability to move up and down. This was a design feature that carried over from the first iteration, however, we have given it a office chair hydraulic cylinder to create the up and down movement. The will be helpful, because the grinder fixture will weigh about 40lbs and the operator will not have to lift that weight themselves. This further satisfies the ergonomic requirements. The grinder will also have two slots to allow the arm to be rotated 90 degrees from over the workbench to over the floor. A magnetic pin will be in the slot to hold the radial motion of the grinder and allow the operator to quickly rotate the grinder arm into the position that is most comfortable.
Force Feedback Subsystem
The objective of this system is to provide the operator instant feedback on the force being applied to the stone tip on the grinder. Through testing, an optimized force will be found to decrease the grinding time and increase the stone tip life. This system will make the grinding process more consistent and faster than it is currently.
The team plans to use an Arduino controller to convert an input signal to force. This force will be compared to the team’s test data. The applied force will be compared to the desired force and that information will be displayed on an LED screen for the operator to get instant feedback.
The initial plan was to use strain gauges to measure the strain in the fixed grinder subsystem. Using the modulus of elasticity, a relationship between the force at the stone tip, and the measured strain was developed. This measurement technique had some difficulties. The fixed grinder system needs be quite stout. Since this will be used in a manufacturing environment, its longevity needs to be ensured. However, if we make the system stout so that there isn’t any deflection, then it will be difficult to accurately measure the needed strain. If the cross-sectional area is reduced to allow for strain, then there is the greater possibility of breaking the system to a force overload from the manufacturing environment.
Since there were several challenges with the initial force feedback system, the team decided to measure the force at the tooling rather than in the grinding fixture system.
Physical Force Feedback
The team decided to pursue a physical force feedback system in lieu of an electrical one. A physical system would have less room for error and would have longevity in the industrial environment. The team was having difficulties in designing the feedback system and thought this change of perspective could provide some useful solutions.
Two main ideas surfaced:
- The first involved a laser pointer that was directed through a concave lens onto a chart. This would allow for feedback from a force in any direction in the tooling plane.
- The next idea was to use two separate torsional spring gauges mounted to the tooling. The operator would use the feedback from both gauges to approximate an appropriate force.
The largest difficulty the team faced with the force feedback system, was that it had to measure a plane of forces, rather than a single direction of force. This meant that most designs would require a system to calculate the resultant force from the two vector directions required to fill the tooling plane.
The physical force feedback systems had two different approaches to this issue:
- The laser pointer design was able to fulfill this requirement because of the symmetry of the rod’s circular cross section. However, power chord for the grinder is in the same space as the rod and we could not move it.
- The feedback design using two separate gauges seemed to be an unnecessary complication for the operators.
Decision to Cut the Force Feedback Subsystem
Over the course of the project the team spent a considerable amount of time considering a force feedback system with electronic sensors, as well a physical feedback designs. After months of design work and consultation with University Professors, the team decided to cut the force feedback system from the project.
There were two main reasons for this:
- The force feedback system did not add enough value to the project to justify the time and resources we were devoting to it.
- The operators would likely be able to listen to the grinding noise and tell if they were applying enough or too much force.
Testing & Validation
Pullout Apparatus Design
For the team to do testing, a pullout contraption had to be created. Boeing uses a hydraulic press that pulls the attached part from below toward the ground. The University of Idaho’s press is only capable of pushing downward. This means that the team had to design a mechanical device that would translate the force from the press from above the tubing, to the below. In addition, they still needed to have the dye and saddle like what Boeing uses. The bottom of the saddle is supported with two blocks from the press. The head of the presses pushes on the top of the design, pulling the dye through the tubing.
Original Testing Plan
All tests will use an identical test sample show below. Each sample is made from an 18” long, 3” 304 stainless steel tubing with a wall thickness of 0.065”. The samples have five holes made of an ellipse, 1.4”x0.8” in size, three on one side and two on the other. All holes are equally spaced. Each hole will be labeled 1-5 with a permanent marker as shown in the diagram below. During the test all holes will be ground evenly for a specified time using the grinding fixture from the project. After the grinding is complete, the digital microscope will be used to photograph the edge of the ductwork on all four corners of the ellipse as shown below. The corners are being selected because the radii create stress concentrations meaning the pullout would most likely tear in those areas. Initial Tests for Grinding Time:
The initial two tests will each use one of the four samples and will be used to set a baseline of both grinding time, and visual feedback from the digital microscope.
The holes in the first test will use the grinding times as follow:
1. No grinding
2. 5 minutes
3. 10 minutes
4. 15 minutes
5. 20 minutes
As stated earlier, each hole will be photographed in its four corners prior to pullout. Each hole is to be pulled into a 2” pullout. It is expected that hole #1, which had no grinding time, will tear during pullout, while hole #5, with 20 minutes of grind time, will not tear.
The grinding times for the second test will depend on which pullouts tore in the previous test. The pullout with the least grinding time that did not tear will give the grinding time for hole 1. Each subsequent hole will add one minute onto the grinding time.
1. Least grinding time from last test that did not fail
2. +1 minutes
3. +2 minutes
4. +3 minutes
5. +4 minutes
Again, the holes must be visually documented prior to pullout. The pullout process will remain the same. It is expected that there will be a cutoff based on grinding time where the pullouts with less grinding time all fail.
Failure of the Original Test Plan
We had two major problems as we did our initial testing. The first problem was that the stone grinder tips that we purchased were far more aggressive than the tips that Boeing was using in their process. This meant that too much material was being removed too fast and wasn't a realistic representation of Boeing's process. We solved this by contacting Boeing and getting the new tips directly from them. The second major problem was that none of our pullouts failed on our test sample. This meant that our tube didn't need to ground to have a successful pullout. This derailed the whole original test plan and we were unable to complete any more testing before the semester ended. A testing plan has been written for our time at Boeing where we will work with the operators to do testing on site. The team did induce cracks by filing the inside edge of the part and were able to create failures. This showed us that the highest stress is in the long side point of the ellipse.
Onsite Test Plan
The grinding tips purchased removed materially much more aggressively than the tips Boeing uses now. Although they may be too aggressive for softer materials like aluminum and stainless steel, these tips may be a great solution to material removal for titanium and Inconel. The surface finish left by the purchased tips was much rougher than the Boeing tips. If the rougher tips are used, they may need to be accompanied by a surface finish step with the Boeing tips.
Test- Select a pullout geometry and material that traditionally takes the operators a significant amount of time, preferably on titanium or Inconel. Use the new tips for the grinding process. With the help of the operator and digital microscope determine if the surface is adequate to pull or if it needs to be smoothed out with a softer tip. Repeat the process with the tips used in production now. Compare grinding times for the old and new tips.
Water Jet Pathing:
When the test samples were cut at WSU’s water jet, the cutter started and stopped on the path of the ellipse. This led to the only natural pullout failure that was achieved in testing.
Test- Ensure that the pathing for the water jet doesn’t start on the ellipse path.
Strain Rate Sensitivity:
Strain rate refers to the speed at which the die is being pulled through the part by the ram. The faster this process happens, the more stress the part will experience internally. While parts at high temperatures experience less stress from the pull, they will have a greater change in internal stresses from a change in strain rate.
Test- Select the heated pull at the highest temperature for each material. Research each material’s strain rate sensitivity at that temperature and determine the degree to which a slower pull would reduce internal stresses. If this yields any significant results, pursue a test with slower pulling (on the order of 20 seconds instead of 2).
All of the team's documents can be found in the Senior Design Folders on the University of Idaho Shared Drive. Take the paths below and paste them into the address bar of your Windows Explorer. The path to the folder is: S:\Engineering\SeniorDesign\- Senior Design\2017-2018\Inconel Tubing PrePullout
S:\Engineering\SeniorDesign\- Senior Design\2017-2018\Inconel Tubing PrePullout\Meeting Minutes
S:\Engineering\SeniorDesign\- Senior Design\2017-2018\Inconel Tubing PrePullout\Presentations
S:\Engineering\SeniorDesign\- Senior Design\2017-2018\Inconel Tubing PrePullout\Final Report
S:\Engineering\SeniorDesign\- Senior Design\2017-2018\Inconel Tubing PrePullout\Test Plans
S:\Engineering\SeniorDesign\- Senior Design\2017-2018\Inconel Tubing PrePullout\Budget and Schedule
|Born and raised in Boise, Idaho, Conor is studying Mechanical Engineering at the University of Idaho. He is set to graduate in May 2018. After graduation Conor will work in industry for a few years before returning to school for a masters in business or engineering.|
|From Idaho Falls, Matt is a senior in Mechanical Engineering at the University of Idaho and is set to graduate in May 2018. He would like to get a graduate degree in engineering and work in the defense industry.|
|Hunter is from Kamiah, Idaho, he is a senior studying Mechanical Engineering at the University of Idaho. Hunter is set to graduate in May of 2018. After he graduates, Hunter would like to work in the aerospace industry.|