MOOG Fixture Leakage Reduction 1
Project Number: P13681
Moog Fixture Leakage Reduction
Austin FrazerMechanical Engineering / Eileen Kobal
Chemical Engineering
Ana Maria Maldonado
Industrial and Systems Engineering / Marie Rohrbaugh
Mechanical Engineering
Abstract
Customer specifications require that Moog, an aerospace company in New York, tests their aerospace parts before they can be used commercially. One of the most commonly used tests is the external leakage test, which makes sure that there are no faults within the part that would cause fluid flowing through it to leak out to the atmosphere.It is desired to reduce the amount of leakage detected by the test through reducing the leakage coming from the fixtures holding the aerospace parts during testing. All the fixtures currently have two O-rings with a vent in-between them to reduce the pressure differential created in the test set-up, and this vent is the specific area of focus for the Multidisciplinary Senior Design (MSD) Project.
This project takes a look at concept ideas proposed by Moog, which are meant to reduce the amount of excess leakage detected when running the tests. The concept went through an iterative process to confirm feasibility and to make sure that other ideas would not outshine it. A prototype pulse-purge system was created to alternately flush the vent with an inert gas and remove gases from the vent. It was built at Rochester Institute of Technology and brought to Moog for testing. Testing results show an increase of gas detected by the external leakage test. While the system did not perform as originally planned, the project gave Moog some great insight as to how their system works and what areas of improvement to keep looking for.
Introduction
Presently, Moog is required by contract to test a variety of customer specifications on spaceflight hardware. Several of these tests are performed in one of two Automated Gas Testers (AGT); one which is depicted in Figure 1. A very common test which Moog desires improved test accuracy is the external leakage test. The basic functionality of the external leakage test in the AGT is given below:
- The valve (or other aerospace part) and various fixturing are assembled in the AGT-can (see Figure 1). Any electrical leads/pneumatic connections are made to the AGT cabinet/computer
- The can is sealed and pumped down to nearly 0 psi by the helium ion detector
- The helium ion detector reads any helium ions that are pulled from the AGT-can.
- A lab technician chooses a test sequence on the computer. The test sequence(s) are completed when the computer triggers an array of valves housed within the AGT cabinet.
Moog believes that some of the helium read by the detector is actually coming from the fixtures that hold the valve in place while testing. Measures have already been taken by Moog to reduce this “fixture leakage” such as the implementation of vented double O-ring seals.
This project aims to reduce the amount of helium leakage in the testing system so that the values read by the helium detector have a higher probability of having originated from the valve itself. The two interfaces of concern are, the double O-ring vent space between the fixture parts acting as a barrier for high pressure helium, and a larger O-ring on the top of the AGT where helium from the atmosphere of the room can pass into the system.
P13681
MOOG Fixture Leakage Reduction 1
Figure 1: Automated Gas Tester used at MOOG
Figure 2: Problem Statement Schematic
P13681
MOOG Fixture Leakage Reduction 1
Background
Before deciding on a possible solution for this problem, some background research was performed to ensure that the design ideas have potential to solve the problem.
The gases readily available at Moog include nitrogen and helium. Knowing that the test gas is helium, the interaction between helium and nitrogen was investigated. The nitrogen molecule is larger than the helium molecule. It was determined that the movement of the gases through a pressure differential would be more effective than the movement of the gas molecules via diffusion. The diffusion of the gases through the O-ring was also investigated. With Moog’s operating conditions, the gases are more likely to travel around the O-ring than through it.
In the early stages of the project, communication with Moog producedthe idea of pulsing the double O-ring vent (found in all the fixtures) with nitrogen gas followed by purging the system with a vacuum. This would fill the vent with a larger molecule gas to encapsulate any of the smaller helium ions, and then pull all of the gases out of the small vent space. Moog had plans to implement an idea like this in the past, but had never tested the concept. Therefore, this concept was included as a possible solution.
A computer model was created to help understand the behavior of gases within the double O-ring vent fixture. Within Simulink, three different cases were modeled. Case one had the pressure in the vent at atmospheric pressure and was used as the baseline for the other models. Case two modeled a constant vacuum on the vent, and case three alternated between pulling a vacuum on the vent and pumping nitrogen into the vent to flush out any helium.Based on the other research, it was assumed that all of the gas flow is pressure driven and that any mixtures of gases are ideal.
Figure 3: Simulink model set-up
Looking at the results for cases one and two, the concentration of helium in the vent dominates the response of the simulation. Case two would show a significant improvement over case one if the percentage of helium was allowed to grow near 100% in both cases and the vent volume was significantly increased. Case three shows a marked improvement over Moog’s original set-up (modeled in case one). This result can be seen in thegraph.
Figure 4: Leakage results from computer simulation
Process
The first step in the process was to understand the scope of the project and what problem the MSD team was to solve. The overall goal was determined to be: reduce the amount of helium leakage from the fixtures in the AGT system during the external leakage test.A functional decomposition took the main goal and split it into smaller manageable “functions” to focus on during the concept generation phase of the project. These functions were: access the gas, move/evacuate the gas, and flush the gas.
After understanding the project, the customer constraints and the engineering specifications needed to be considered.This testing system is operated in the clean room, so all of the products needed to comply with the cleanliness
P13681
MOOG Fixture Leakage Reduction 1
certification. The shape of the testing fixture could not be modified by the MSD team, and all designs must be able to attach to the AGT system. The biggest point stressed by the customer was that any designs must be safe for the operators to use. A comparison of all the customer needs and the engineering specifications can be seen in Figure 5. The numbers in the figure represent how well the specifications meet the requirements.
It was decided that this project would be implemented at Moog in East Aurora, NY. The resources available to the MSD team at Moog included nitrogen gas up to 120 psia, and a vacuum system of 1 psia within the clean room. Moog resources also included contact with operators who work with the AGT system and Moog engineers.
Figure 5: Customer specifications and requirements
P13681
MOOG Fixture Leakage Reduction 1
Once anunderstanding of the scope and all the parameters was reached, a few concept ideas were generated for each functionality needed to complete the project. The concept ideas were put into a Pugh matrix so that they could be compared with each other against a list of criteria generated from the project specifications.
Table 1: Pugh matrix comparing ideas for Move/ Evacuate the gas
Included in these concepts was the original pulse of nitrogen and purge with a vacuum in the small double O-ring vent space proposed by Moog. The concepts that “passed” the Pugh matrix were run through the aforementioned Simulink model on the computer for concept validation.
The concepts chosen include a system at the helium inlet that would alternate between purging the vent with a vacuum and filling the vent with nitrogen gas to flush it out, as originally proposed by Moog. This concept was the only one that fit within all the constraints given by the customer. Another, simpler system would flush the existing large O-ring at the top of the testing can with nitrogen to act as a barrier between the testing can and the helium in the atmosphere. Both of these systems would be incorporated within the existing test sequence and operated by a LabView program.
After concepts were chosen, models of the final design needed to be created. These models were designed in Creo Parametric.Designs included a plug to attach the nitrogen/ vacuum system to the helium inlet on the AGT and also layouts of how the tubing, valves, and wiring would need to be in order to operate the system. A diagram of how the LabView would need to interact with the gas systems was also generated. Finalized models were sent to Moog so that any machining could be done to clean room standards.
A battery of tests was produced to be certain that the final design would meet all of the necessary specifications. The table below lists the testing requirements.
Table 2: Testing to be done based off of specifications
Spec. # / Function / Test / Nominal / Pass/Fail / Units1 / Reduction of Test gas Leakage / Run external leakage test at Moog with their current system using a blank instead of a valve and new O-rings. Run external leakage test with the new system under the same conditions. Verify the reduction percentage of helium comparing both results. Redo both tests using used O-rings. / 10% / ±1% / cm3/sec
2 / Amount of Nitrogen / Measure the amount of nitrogen flowing into the system using a flow meter while the external leakage test is in operation with the new system / <100 / N/A / scc/min
3 / Pressure Condition / Each part model will be run through finite element analysis to verify that all parts can work together under this pressure / <3500 / N/A / psi
4 / Cost / Create Bill Of Materials and verify that the Total cost for one System doesn't exceed the budget / N/A / $
This project took a break over the winter quarter so that any machining done at Moog. Unfortunately one of the risks discussed in the planning phase actually happened. The funding at Moog was cut, so therefore the funding for the project was cut. This meant that designs needed to be shifted so that they could fit within the new project budget. Possible alterations included designing a prototype system that would be manufactured and tested at RIT. The other idea was to create a less sophisticated system that would be tested at Moog, but would not be used for production parts.
It was decided that a nitrogen gas/ vacuum system for the vent would be created and tested at Moog. Fewer constraints on the system allowed the team to create the system with one computer controlled valve and the others controlled manually. With the shortened timeline and budget, the original plans to permanently alter the AGT system for the large O-ring flush system were cancelled. In the primary designs, all of the LabView would be programed by Moog’s LabView vendor. Due to cost reductions, the MSD team took over the necessary programming in order to connect the system to a computer. The team then went forward with building necessary components and purchasing parts. Many of the parts used for the project were spare parts at Moog that could be used for the project and then easily taken apart later on if needed by someone at Moog.
Final Setup
P13681
MOOG Fixture Leakage Reduction 1
Figure 6: Diagram of the final set-up
Our system design included a pulse purge nitrogen system, where nitrogen is pumped into the vent, and then pulled out by a vacuum. This cycle is controlled by a three-way valve with two inlets and an outlet. A check valve is put in line with the vacuum to protect the vacuum from any extreme pressure introduced by misplaced O-rings. This risk is very common, and the test procedures include damaged seals. A relief valve is added to the system so that the fittings are not under any pressure differential at disassembly. A pressure regulator is implemented, so that the nitrogen pressure can be varied for optimization.
The 3-way valve in the system is controlled by LabView. The time for each step can be changed on the main screen for optimization purposes. With the Simulink model as a guideline for how the system should run, the test plans for this optimization were developed.
P13681
MOOG Fixture Leakage Reduction 1
In order to test the system, we needed to interface with the existing system. We redesigned the gas inlet plugs used in the system to fit our needs. The Helium inlet had a ¼” tube welded to it to represent a valve with a ¼” tube inlet. The other end of the tube went into the existing fixturing for this type of unit, and the pressure was stopped at the plug where the double O-ring with vent configuration is present. The vent is connected to the system described above.
Standard sizing for all fittings and couplings have been selected (mostly ¼ AN/pipe). The plugs have been machined to interface with ¼ inch AN fittings. Very little initial cost (<$50 not including borrowed components) was required to assemble and test the proof of concept system. The permanent subsystem would require some additional maintenance such as the occasional cleaning/changing of seats of the circle seal valves. This is fairly trivial given the current amount circle seal valves that already are subjected to scheduled conditioning in the clean room.
Components all have pressure ratings higher than the pressures being introduced to our system. Everything is bolted down to the plastic baseplate, so that everything is completely fixed when the system is pressurized. Maximum pressure ratings for any “proof-of-concept system" components were not exceeded at any point during data collection. No failures were observed during test.
Results and Discussion
The MSD system was taken to Moog to be tested in the clean room with the AGT system. Every part was cleaned to clean room standards prior to testing. Three main cases were tested: a pulse of nitrogen with a purge from the vacuum source, a constant flow of nitrogen, and a constant application of the vacuum.
P13681
MOOG Fixture Leakage Reduction 1
Figure 7: Helium leakage vs. Helium pressure with Pulse-Purge system
Figure 8: Leakage per second showing 14 cycles of pulse-purge system
P13681
MOOG Fixture Leakage Reduction 1
The pulse purge system provides very noisy data averaging at much higher leakage rates than the baseline. The duty cycle does not present any significant relation to the leakage.The leakage through the o-rings proves to spike when pressure is introduced to the vent between the o-rings. At small nitrogen pulses and long vacuum cycles, the trend becomes obvious. Contrary to the simulated predictions, as the pressure of nitrogen is increased, the helium leakage is also increased. This trend is more apparent with higher pressures.
P13681
MOOG Fixture Leakage Reduction 1
Figure 9: Helium leakage vs. Helium pressure over a constant supply of nitrogen
Figure 10: Helium leakage vs. Helium pressure over a constant vacuum
P13681
MOOG Fixture Leakage Reduction 1
As well as testing the pulse purge system, the system can be tested with constant pressure, and constant vacuum. With constant pressure, the average leakage trended above the baseline. With vacuum, no significant change in leakage occurred. Both had no significant trends, seen in Figures 9 and 10.
With the data from the test results, a regression analysis was performed, using MiniTab, to see if there might be any specific correlations. All the testing situations were combined in three main circumstances: just nitrogen, just vacuum, and pulse of nitrogen/ purge with vacuum. The dependent variable in the regression analysis was the amount of helium leakage detected.