1 Lockheed Martin: Redesign a Heat Exchanger using additive manufacturing methods 5 May 2016 EDSGN 100 Section 026 Kayla Badamo, Daniel Jackson, Josh Lutton SUMMARY: Lockheed Martin needed a solution that would improve the design of their heat exchanger through the use of additive manufacturing. Lockheed Martin required that the redesigned heat exchanger had the same dimensions as the original, better or the same amount of heat transfer, and could be created using additive manufacturing. Research was conducted into heat exchanger designs and materials. Three prototypes were created in SolidWorks and then printed in PLA using fusion deposition modeling. These prototypes were then tested in order to determine the optimum design. Following the engineering design process, a successful final design was produced.
2 Table of Contents Section Page Introduction 2-3 Methods 3-4 Results and Discussions I 4-6 Results and Discussions II 6-7 Conclusions 7-8 Appendices 8-10 References 11 I. Introduction The project presented by Lockheed Martin was to redesign a heat exchanger for production with additive manufacturing. This heat exchanger would most likely be used on one of Lockheed s aircrafts. The current process used to produce a heat exchanger requires significant labor and costs using traditional subtractive manufacturing methods. The primary goal was to create a more efficient heat exchanger without changing the heat exchanger s dimensions. To optimize the heat exchanger, the team needed to research heat exchangers in order to understand how they operate. A heat exchanger works to remove heat from the circuit board by using metal fins. The fins absorb the heat of the circuit board and air flowing through the fins removes the heat and takes it out of the system. In order to create a more efficient heat exchanger two aspects must be considered, surface area and thermal conductivity. Surface area is critical since larger amounts of metal means more heat can be absorbed at once. A material with good thermal conductivity is necessary for rapid absorption and removal of heat. After extensive research, the solution to
3 creating a more efficient heat exchanger is to increase the surface area using a lattice structure that, while unable to be produced using subtractive methods, is feasible using additive methods that utilize a material with a good thermal conductivity, such as aluminum. (Concord.org). II. Methods Three solutions were considered for the redesign of the heat exchanger; they were large two-way hexagons lattice structure, small one-way hexagons lattice structure, and small two-way hexagons lattice structure (Figure 1). The lattice structure should increase the surface area while maintaining an unimpeded airflow which is the reasoning behind all three solutions being latticelike structures. Six criteria were decided upon for the solution requirements: increased surface area, same or improved airflow, approximate weight as original, same size as original, and reduced cost and build time. Increased surface area was the most important criteria since that had the greatest effect on optimizing the heat exchanger, which is why it was weighted most highly in the selection matrix (Table 1). Same or improved airflow and same size as original were weighted the same; airflow needs to be maintained so that the heat exchanger can be cooled easily and dimensions must stay the same size since it has to fit within a small space in an aircraft s circuitry. Approximately the same weight was weighted next highest weight since the heat exchanger is on an aircraft and the weight cannot change drastically or else it could affect the flight of the aircraft. Cost and build time were weighted the least because although they are still essential, they can be sacrificed if necessary for significant optimizations to the heat exchanger s overall design. To evaluate the three prototypes four tests were conducted. The four tests were a water flow test, airflow test, surface area test, and weight test. The water flow test simulates how steadily
4 the flow is through the exchanger and where it is directed. Water will be poured through the exchanger and if the water deviates from its original stream too much throughout the exchanger the solution passes the test. The airflow test simulates how well air flows through the exchanger. Compressed air will be blown through the lattice structure and hit against a piece of paper behind the exchanger. If air flows through the exchanger and blows the paper backwards a significant amount, the solution passes the test. The surface area and weight test will be conducted through SolidWorks analysis. The material of the heat exchanger is changed to aluminum, the material exchanger will be manufactured from, and SolidWorks is able to calculate the surface area and weight of the exchanger. If the surface area and weight of the solution are greater than the surface area and weight of the original exchanger then the solution passes the test. III. Results and Discussions I After evaluating the three initial solutions, model 3, the small, two way hexagonal design was chosen as the superior solution (Table 1). Model 1 scored poorly in the selection matrix due to poor surface area and weight. This solution lost surface area when calculated in solidworks, so it was given a rating of one out of five for surface area (Table 1). Model 1 also rated at a three out of five when its weight was calculated to be heavier than that of the original (Table 1). Model 2 scored significantly better than model 1 thanks to the simplicity of the design. The surface area calculated in solidworks increased over the original design, but not significantly, resulting in a three out of five rating (Table 1). However, in all other criteria, model 2 scored high with a four out of five for airflow and five out of five ratings for all other criteria (Table 1). Model 3 accumulated the largest score in the solution matrix. Model 3 actually scored the lowest of the
5 three models in the airflow tests with a three out of five rating (Table 1). Despite the low airflow rating, model 3 gained significant surface area over the other models and the original design. When designing the prototype, the biggest limitation on what was possible was the capabilities of additive manufacturing. To successfully create a printed prototype, the geometry of the model needed to be within the physical limitations of additive manufacturing. To accommodate these limitations, hexagons were chosen to replace the fins of the original heat exchanger. For PLA extrusion, the maximum, unsupported angle that can be extruded is 45 degrees from the vertical. This made hexagons the ideal shape since the unsupported angles would be 60 degrees. Spacing between each row of hexagons was also important. The spacing needed to be wide enough to allow for the airflow of the intersecting rows of hexagons, but narrow enough to ensure that valuable surface area was not lost. Build time was also taken into account since a full scale print takes over a day to print. So the prototype printed was a quarter section of the complete model. Printing only a section of the full model meant that the prototype could be printed in several hours, and still be capable of performing in the tests. Once the prototype was printed, testing could begin. In the airflow test, the prototype gave encouraging results. Airflow was somewhat hindered because of the intersecting rows of hexagons, but it appeared that the spacing between each row minimized this restriction. Similar results were yielded in the water flow test. In this test, the stream of water was diffused throughout the prototype with minimal restriction to the flow of water. Both the surface area test, and the weight test were completed in SolidWorks. Weight was significantly heavier than that of the original design, but the surface area was also significantly greater than that of the original. After conducting the test, several conclusions were made about how to improve on this design. It was determined that the spacing in between the rows of
6 hexagons could be increased slightly. This change would allow for improved airflow, but would have little impact on the surface area. IV. Results and Discussions II A second round of prototypes was never created due to time limitations; however, additional research was done to see how the initial prototype could be improved, especially considering the additional manufacturing constraints of metal additive. One of the more important details learned was that additive manufacturing with metal sometimes requires a more restricted geometry. This is the case for powder bed fusion, the maximum, unsupported angle possible is 60 degrees; additionally, there can be no bridging in metal additive (Additive Manufacturing:... ). This meant that if powder bed fusions was to be used, the current design would not be possible at the current orientation. To overcome this issue, the design could be rotated in such a way that eliminated bridging and kept every angle above 60 degrees or the hexagons could be replaced with diamond patterns. This pattern would be possible if using powder bed fusion, but it could complicate the intersecting pattern idea that was used with the hexagons. If second round prototypes had been created, it is predicted that it would have similar characteristics to the first prototype, but exploring different geometries. No successful 3D scans were made because the prototype only altered the design inside the housing of the exchanger, and it was difficult to capture these internal features with the 3D capabilities available. Compared to the original heat exchanger, this prototype greatly improved surface area thanks to additive manufacturing. Extruding material to create the heat exchanger allowed for an intricate design to be used inside the housing. This is something that cannot be done with traditional subtractive methods, and also created a greater amount of surface area.
7 If Lockheed Martin were to use this design to replace their current design, aluminum powder bed fusion is the recommended process. Aluminum is affordable, lightweight, and a proficient conductor of heat. Using aluminum also means that it is possible to extrude this using the powder bed fusion process of manufacturing, which was determined to be the most effective process for metal extrusion. Production of this design would be limited to the accessibility to a powder bed fusion machine. Thankfully, Lockheed Martin already has additive manufacturing capabilities, so the part could be created on site where the aircrafts are produced. This would save Lockheed Martin both transportation time and costs. Expected cost is high since the aluminum powder must be very fine and precise. Lockheed Martin can expect the cost of an individual piece to be in the area of $8,550 (Table 2). It is also expected that the build time will be close to 6.3 days for a single piece (Table 3). However, multiple pieces could be printed at the same time thanks to the capabilities of powder bed fusion. V. Conclusion Lockheed Martin presented five products that can be improved using additive manufacturing. The team chose to redesign a heat exchanger for production with additive manufacturing. The subtractive manufacturing methods to make the heat exchanger were expensive, had a long build time, and limited geometries. The goal was to make a more efficient exchanger utilizing the capabilities additive manufacturing in order to improve the exchanger and reduce cost and build time. To optimize the heat exchanger, surface area and thermal conductivity of the material had to increase (Concord.org). The team decided to use a hexagonal lattice structure that would significantly increase the surface area. The team also decided to use aluminum as the material because it has a high thermal conductivity and it can be used in additive manufacturing. The most positive design was the two-way hexagonal lattice structure
8 that tripled the surface while keeping a good airflow. To create this exchanger, the material used should be aluminum because it has a high thermal conductivity. The additive manufacturing method that should be used is powder bed fusion (direct metal laser sintering) because a aluminum powder can be used to make the exchanger. The main lessons the team learned from this project were how to work as a team, the different methods of additive manufacturing, and getting more experienced with SolidWorks. VI. Appendices Figure 1- SolidWorks Solutions (Left to right: large two-way hexagons, small one-way hexagons, small two-way hexagons). These images are the CAD renderings of our three prototypes.
9 Figure 2-3D Printed Models, ¾ view. (Left to right: model 1; large two-way hexagons, model 2; small one-way hexagons, model 3; small two-way hexagons). These were printed with extruded PLA. Table 1: Design Selection Matrix This table displays the results of a criteria weighting to choose the best solution.
10 Table 2: Cost Analysis This table displays the cost to additively manufacture the exchanger using direct metal laser sintering (Shapeways.com). Table 3: Build Time Analysis This table displays the build time to create the heat exchanger using direct metal laser sintering (Dmlstechnology.com).
11 VII. References "Additive Manufacturing: Opportunities and Constraints". Rep. Royal Academy of Engineering, 23 May 2013. Web. 25 Apr. 2016. Concord.org. "Heat Transfer." AccessScience (n.d.): n. pag. Concord. The Concord Consortium. Web. Dmlstechnology.com. "DMLS Machines." DMLS Machines. DMLS Technology, n.d. Web. 26 Apr. 2016. Shapeways.com. "Aluminum 3D Printing Material Information - Shapeways." Shapeways.com. Shapeways Inc., n.d. Web. 26 Apr. 2016.