NEW DIAGNOSTIC METHODS FOR THE ANALYSIS OF PRODUCTS CREATED BY ADDITIVE MANUFACTURING

Davide Mavillonio - Asotech Srl
Fabrizio Rosi - TEC Eurolab Srl
Fabio Esposito - TEC Eurolab Srl
NEW DIAGNOSTIC METHODS FOR THE ANALYSIS OF PRODUCTS CREATED BY ADDITIVE MANUFACTURING

Introduction

In modern industry the reliability issues of mechanical components play an important role: industrial computed tomography coupled (iCT) with FEM structural analyses makes it possible to perform fatigue testing to improve the acceptability criteria of molded or additive manufactured components.

Asotech and TEC Eurolab have combined their expertise to create a new service that makes it possible to identify defects in molds using tomography and to perform FEM fatigue testing of the identified geometry; it is also possible to expand the analysis to include experimental validation and optimization tests for the component.

In order to illustrate in detail the potential of the new service, industrial tomography and FEM structural analyses were applied to defects in aircraft propeller blades made using additive manufacturing.

DEFECTOLOGY and TOMOGRAPHY

The world of additive manufacturing exhibits some typical technology defects: TEC Eurolab, with the assistance of industrial computed tomography and metallurgical analysis, has developed a large database over the years (as seen in Figure 1); the most common problems are the inclusion of gas and the lack of fusion.

Computerized Industrial Tomography is a radiographic inspection method that gives information about a 3D object starting from a certain number of radiographic projections made on cross-sections or the complete volume of an object. These scans include the actual shape and any internal defects with resolutions that can go to below 10μm. TEC Eurolab has two tomographic systems that quickly process large components such as crankcase engine or turbine blades. The first one is a NSI X5000 with a 240kV Micro-focus X-ray tube that makes it possible to scan volumes of cylinders up to 1300 mm in height and 500 mm in diameter with resolutions which go to below 10 μm for small volumes.  The second one is a NSI X7000 with a 450kV Mini-focus radiogenic tube that scans volumes of cylinders up to 3000 mm in height and 1500 mm in diameter with resolutions greater than 100 μm using special scanning modes.
Fig. 1 - Defects detectable in additive manufacturing and molds.

INDUSTRIAL COMPUTED TOMOGRAPHY and VIRTUAL ANALYSES

With the precision achieved by the iCT on virtual models of exact size and containing any defects produced by the various production phases (whether traditional or innovative, such as additive manufacturing) it is possible to extend the potential of finite element analysis to also include the actual shapes of finished parts. The STL format model can thus be subjected to FEM testing in order to investigate the possible negative consequences of the defects present or differences in size. The correlation between virtual analyses and experimentation can then be extended by performing tensile tests or fatigue tests on the components analyzed to validate the results obtained from the FEM fatigue testing. The FEM tests that can be performed with this technique are the same that can be obtained using 3D models generated by traditional CAD tools; the difference in this case, however, is the availability of the actual geometry of the component, including manufacturing defects and imperfections. All of this offers the ability to investigate during quality control checks whether a defect (recurring or not) might generate unacceptable reliability in components subject to fatigue. The proposed procedure is thus useful for increasing the reliability of mechanical components and generally offers an innovative approach to the issue of quality control (Chart, Figure 2).

TEC Eurolab and Asotech have begun a collaboration in which the former offers its expertise in the field of tomography, in the implementation of STL models and experimental testing, and the latter provides its expertise in FEM analysis and design for the resolution of critical issues.
 
Fig.2 - Logic flow of the tomography and FEM testing procedure.

AERONAUTICAL TURBINES: case study

To demonstrate the potential of the method developed, it was decided to produce samples of an AlSi10Mg aluminum aeronautical turbine blade made with additive manufacturing (Figure 3) containing defects typical of the technology: gas porosity and non-fusion (two shapes containing the latter). As shown in Figure 4, the printed defects introduced directly into the CAD model used for the printing of the STL model and submitted to iCT, were detected by the tomography and transformed into a STL format file.
Fig. 3 - Printed shape.
Fig. 4 - Printed shape and defects.
The tomographic scan reveals the defects introduced; in particular, the shape with the porosity defect shows the maintenance of the spherical shape (Ø0,8 mm) with partial collapse of the printed geometry. (Figures 5 and 6)
Fig. 5 - Tomography results
Fig. 6 - Tomography results

BLADE FATIGUE TEST

Starting from the STL files generated by the tomography, Asotech performed a fatigue test on the blade in order to establish the life of the component: given the desire to make a correlation between the virtual and experimental results, it was decided to use a simple alternating tensile/compression setup (Figure 7).

 
Figura 7 - FEM model setup in Ansys Workbench.
During the test, particular attention was paid to the meshing of the models, with the introduction of more refined elements close to the defect and the fillet radii on the bars; in fact, these are the areas most prone to potential fatigue failures. As regards the material, an isotropic model was used with characteristics typical of aluminum and fatigue life of 2,000,000 cycles with an amplitude of 100 MPa. As regards the defect due to gas, an alternating force of 10,000 N applied to the bar was used. The FEM testing shows a marked criticality both on the fillet radius and in the area with the defect; in the area of the joint, the stresses appear to be very widespread and homogeneous and there are no singularities: with the force value indicated, the estimated fatigue life of the fillet is about 680,000 cycles. In the defect area, on the other hand, border effects are evident on the irregular zones of the mesh (Figure 8);
Fig. 8 - Mesh used for the analysis
It is conceivable that the component will yield locally in these zones, but it will be difficult for the crack to spread.
The results obtained for the other geometries analyzed are very similar. In particular, the blade with the non-fusion issue, excited by an alternating force of 7,500 N, showed a possible fatigue failure on the fillet radius after 1,380,000 cycles. On the other hand, the analyses of the latter defect show criticality after 1,500,000 cycles with a force of 6,750 N (Figure 9).
Fig. 9 - Fatigue safety coefficient detected in the FEM analysis.
Fig. 10 - Fatigue safety coefficient detected in the FEM analysis for other geometries.
Fig. 11 - Fatigue safety coefficient profile.
The graph shows the fatigue life profile of the various components in the comparison between the real and the theoretical geometry. We can see that the introduction of the use of the STLs allowed on one hand better correlation of the data and, on the other hand, has made it possible to correct the estimated fatigue life with values on average of less than 30% (Figure 11).
 

FATIGUE TESTS

In order to validate the results obtained from the FEM analyses, fatigue tests were performed using a Zwick Roell HB500 servo-hydraulic dynamic testing system (Figure 12) at TEC Eurolab:
Fig. 12 - Fatigue test machine.
The device enables fatigue tests with a frequency up to 30 Hz, loads up to 500 kN and specimens with dimensions up to 1000 mm and mass of 250 kg. The tests showed an excellent correlation between experimentation and virtual reality; the proposed method is thus robust and offers the ability to introduce real defects into the FEM analysis and obtain excellent correspondences with experimental tests (Table 1).
Tabella 1 - Results obtained.
Fig. 13 - Metallographic analysis near the defect.
After the experimental tests, metallographic analyses (Figure 13) showed that, as indicated by the FEM analysis, the crack did not spread in the area of ​​the defect. Furthermore, some porosity from gas close to the defect area was evident, which, however, did not affect the structural resistance of the component.
SUBSCRIBE TO THE NEWSLETTER

Tag:  additive manufacturing analysisadditive manufacturing concept processadditive manufacturing design industrial processadditive manufacturing laboratoryadditive manufacturing material selectionadditive manufacturing metalsadditive manufacturing plasticsadditive manufacturing standardsAdditive Manufacturing TestingAdditive Manufacturing Testing laboratoryadditive manufacturing testing processadditive manufacturing testing serviceswhat materials can be used in additive manufacturing?

Sigla.com - Internet Partner
Contact us to learn more or request your tomographic analysis click here X